# AIHubMix — all verified model integration guides (llms-full.txt) > One-fetch concatenation of every verified per-model guide (415 models). Canonical per-model form: any model page plus `/llms.txt`, e.g. https://aihubmix.com/model/gemini-omni-1.2-flash/llms.txt — site index: https://aihubmix.com/llms.txt — agent onboarding: https://aihubmix.com/agents.md Models without a verified guide are not included; discover them in the live catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order # Gemini Omni 1.2 Flash Model id on AIHubMix: `gemini-omni-1.2-flash` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Gemini Omni Flash (gemini-omni-1.2-flash) is a high-performance multimodal model designed for high-speed video generation, editing, and cinematic-level control. It can generate or edit videos up to 15 seconds long with native audio from text, images, and video, and better maintain consistency of people and subjects. - Developer: Google - Input modalities: text, image, video - Release date: 2026-11-07 - Pricing: per-generation — see https://aihubmix.com/model/gemini-omni-1.2-flash ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /ai/v1/videos/{id}` until a terminal state ## Example ```bash curl -s https://aihubmix.com/ai/v1/videos \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-omni-1.2-flash","prompt":""}' ``` The response contains a task id — poll `GET https://aihubmix.com/ai/v1/videos/` until `status` is terminal, then read the output URL(s). Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (video endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string|null | no | "16:9" | | | `duration` | null | no | | | | `extra` | object,null | no | | | | `frame_images` | array,null | no | | | | `input_references` | array,null | no | | | | `model` | string | yes | | | | `prompt` | string | yes | | | | `resolution` | string|null | no | | | | `seed` | integer,null | no | | | | `webhook_url` | string,null | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/gemini-omni-1.2-flash/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/gemini-omni-1.2-flash/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/gemini-omni-1.2-flash - Try in browser: https://aihubmix.com/playground?model=gemini-omni-1.2-flash - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-omni-1.2-flash/llms.txt --- # GPT 6.1 Terra Model id on AIHubMix: `gpt-6.1-terra` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-6.1 Terra is OpenAI’s multimodal reasoning model, supporting text and image inputs for tasks such as product photo matching, technical drawing interpretation, application interface interaction, and text drafting and revision. It combines speed and efficiency with visual reasoning, precise computer use, and clear, natural writing. - Developer: OpenAI - Context window: 1,050,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Computer use, File search, Structured outputs, Prompt caching - Release date: 2026-10-26 - Pricing: $1/M input tokens, $5/M output tokens, $0.05/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-6.1-terra","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-6.1-terra","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-6.1-terra - Try in browser: https://aihubmix.com/playground?model=gpt-6.1-terra - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-6.1-terra.546abdd3.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-6.1-terra/llms.txt --- # GPT Image 2.6 Flare Model id on AIHubMix: `gpt-image-2.6-flare` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-Image-2.6 Flare is OpenAI’s image generation and editing model for posters, product imagery, and rapid creative iteration. The Images 2.6 family delivers 50% faster image generation, with Flare producing images in as little as six seconds at low quality and improving layout quality for professional visuals. It supports GIF creation, SVG exports, markup-guided editing, and a complexity parameter for controlling scene planning. - Developer: OpenAI - Input modalities: text, image - Release date: 2026-11-05 - Pricing: per-generation — see https://aihubmix.com/model/gpt-image-2.6-flare ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/edits` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-image-2.6-flare"}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `async` | boolean | no | false | | | `complexity` | string,null | no | | low, medium, high, null | | `extra` | object,null | no | | | | `frame_count` | integer,null | no | | | | `frame_duration_ms` | integer,null | no | | | | `image` | - | no | | | | `images` | array,null | no | | | | `markup` | string | no | | | | `mask` | string | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `output_format` | string,null | no | "png" | | | `prompt` | string | no | | | | `quality` | string,null | no | | low, medium, high, xhigh, max, auto, … | | `size` | string|string|null | no | | | | `svg` | object | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/gpt-image-2.6-flare/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/gpt-image-2.6-flare/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/gpt-image-2.6-flare - Try in browser: https://aihubmix.com/playground?model=gpt-image-2.6-flare - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-image-2.6-flare/llms.txt --- # GPT Image 2.6 Sunburst Model id on AIHubMix: `gpt-image-2.6-sunburst` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-Image-2.6 Sunburst is OpenAI’s image generation and editing model for visual work requiring detail, precision, and consistency across edits. The Images 2.6 family delivers 50% faster image generation, while Sunburst offers higher-quality generation and greater consistency at medium or high quality settings. It supports GIF creation, SVG exports, markup-guided editing, and a complexity parameter for controlling scene planning. - Developer: OpenAI - Input modalities: text, image - Release date: 2026-11-05 - Pricing: per-generation — see https://aihubmix.com/model/gpt-image-2.6-sunburst ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/edits` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-image-2.6-sunburst"}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `async` | boolean | no | false | | | `complexity` | string,null | no | | low, medium, high, null | | `extra` | object,null | no | | | | `frame_count` | integer,null | no | | | | `frame_duration_ms` | integer,null | no | | | | `image` | - | no | | | | `images` | array,null | no | | | | `markup` | string | no | | | | `mask` | string | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `output_format` | string,null | no | "png" | | | `prompt` | string | no | | | | `quality` | string,null | no | | low, medium, high, xhigh, max, auto, … | | `size` | string|string|null | no | | | | `svg` | object | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/gpt-image-2.6-sunburst/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/gpt-image-2.6-sunburst/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/gpt-image-2.6-sunburst - Try in browser: https://aihubmix.com/playground?model=gpt-image-2.6-sunburst - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-image-2.6-sunburst/llms.txt --- # Hy4 Model id on AIHubMix: `hy4` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Tencent Hy4 preview is a Mixture-of-Experts language model designed for coding, office analysis, and scientific research. With a 1M-token context window, it supports complex software development, long-document analysis, and multi-step workflows. Compared with Hy3 preview, it expands model scale and context capacity while strengthening performance on practical productivity tasks. - Developer: Hunyuan - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2026-11-04 - Pricing: $0.8823/M input tokens, $2.647/M output tokens, $0.4412/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"hy4","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"hy4","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/hy4 - Try in browser: https://aihubmix.com/playground?model=hy4 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/hy4.be4de3a6.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/hy4/llms.txt --- # Kimi K3.1 Pro Model id on AIHubMix: `kimi-k3.1-pro` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Kimi K3.1 Pro is based on Moonshot AI’s Kimi K3.1, offering code generation and interactive content development capabilities for coding assistance, code analysis, and visual prototyping. - Developer: Moonshot AI - Context window: 1,048,576 tokens - Input modalities: text, image, video - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Release date: 2026-10-24 - Pricing: $2/M input tokens, $10/M output tokens, $0.2/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"kimi-k3.1-pro","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"kimi-k3.1-pro","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/kimi-k3.1-pro - Try in browser: https://aihubmix.com/playground?model=kimi-k3.1-pro - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/kimi-k3.1-pro.6fc0ce5f.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/kimi-k3.1-pro/llms.txt --- # Kimi K3.1 Pro Highspeed Model id on AIHubMix: `kimi-k3.1-pro-highspeed` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Kimi K3.1 Pro Highspeed is an accelerated serving option for Kimi K3.1 Pro, focused on responsive code generation for interactive coding, rapid prototyping, and latency-sensitive applications. - Developer: Moonshot AI - Context window: 1,048,576 tokens - Input modalities: text, image, video - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Release date: 2026-10-24 - Pricing: $4/M input tokens, $21/M output tokens, $0.2/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"kimi-k3.1-pro-highspeed","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"kimi-k3.1-pro-highspeed","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/kimi-k3.1-pro-highspeed - Try in browser: https://aihubmix.com/playground?model=kimi-k3.1-pro-highspeed - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/kimi-k3.1-pro-highspeed.e1fc5640.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/kimi-k3.1-pro-highspeed/llms.txt --- # Crouton Alpha (free) Model id on AIHubMix: `crouton-alpha-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Developed by Stealth, crouton-alpha-free is a lightweight model built for scalable coding, agentic, and multimodal workflows. Featuring an expansive context window of 1,048,576 tokens, it is ideal for executing agentic sub-tasks, tool use, and deep document analysis. This model is highly optimized for developers looking to streamline high-volume, repetitive tasks. - Developer: Stealth - Context window: 1,048,576 tokens - Input modalities: text, image, video, document, audio - Capabilities: Thinking, Tool calling, Structured outputs - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"crouton-alpha-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"crouton-alpha-free","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/crouton-alpha-free - Try in browser: https://aihubmix.com/playground?model=crouton-alpha-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/crouton-alpha-free.8f94fd9b.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/crouton-alpha-free/llms.txt --- # Ideogram Vector 1 Model id on AIHubMix: `ideogram-vector-1` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Ideogram Vector 1.0 is Ideogram’s first AI vector (SVG) generation model. It can directly generate high-quality, editable vector graphics from text prompts or existing images. - Developer: Ideogram - Input modalities: text, image - Release date: 2026-10-22 - Pricing: per-generation — see https://aihubmix.com/model/ideogram-vector-1 ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"ideogram-vector-1"}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string|string|null | no | "auto" | | | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `image` | - | no | | | | `images` | array,null | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `prompt` | string | no | | | | `quality` | string,null | no | "medium" | low, medium, high | | `response_format` | string,null | no | "url" | url, null | | `seed` | integer,null | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/ideogram-vector-1/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/ideogram-vector-1/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/ideogram-vector-1 - Try in browser: https://aihubmix.com/playground?model=ideogram-vector-1 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/ideogram-vector-1/llms.txt --- # DeepSeek V4.1 Flash 1017 Model id on AIHubMix: `deepseek-v4.1-flash-1017` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > The official latest version of DeepSeek V4.1 Flash has been upgraded to DeepSeek-V4.1-Flash-1017. API usage and pricing remain unchanged—you can still access the latest model using deepseek-flash. - Developer: DeepSeek - Context window: 1,000,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2026-10-17 - Pricing: $0.155/M input tokens, $0.62/M output tokens, $0.0031/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"deepseek-v4.1-flash-1017","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"deepseek-v4.1-flash-1017","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/deepseek-v4.1-flash-1017 - Try in browser: https://aihubmix.com/playground?model=deepseek-v4.1-flash-1017 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/deepseek-v4.1-flash-1017.5984a577.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/deepseek-v4.1-flash-1017/llms.txt --- # Gemini Omni 1.1 Flash Model id on AIHubMix: `gemini-omni-1.1-flash` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Gemini Omni Flash (gemini-omni-1.1-flash) is a high-performance multimodal model designed for high-speed video generation, editing, and cinematic-level control. - Developer: Google - Input modalities: text, image, audio, video - Release date: 2026-08-28 - Pricing: per-generation — see https://aihubmix.com/model/gemini-omni-1.1-flash ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /ai/v1/videos/{id}` until a terminal state ## Example ```bash curl -s https://aihubmix.com/ai/v1/videos \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-omni-1.1-flash","prompt":""}' ``` The response contains a task id — poll `GET https://aihubmix.com/ai/v1/videos/` until `status` is terminal, then read the output URL(s). Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (video endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string|null | no | "16:9" | | | `duration` | null | no | | | | `extra` | object,null | no | | | | `frame_images` | array,null | no | | | | `input_references` | array,null | no | | | | `model` | string | yes | | | | `prompt` | string | yes | | | | `resolution` | string|null | no | | | | `seed` | integer,null | no | | | | `webhook_url` | string,null | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/gemini-omni-1.1-flash/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/gemini-omni-1.1-flash/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/gemini-omni-1.1-flash - Try in browser: https://aihubmix.com/playground?model=gemini-omni-1.1-flash - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-omni-1.1-flash/llms.txt --- # GLM 5.3 Prime Model id on AIHubMix: `glm-5.3-prime` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GLM-5.3-Prime is the high-speed version of GLM-5.3, with comprehensive advancements in programming and agent capabilities that reach open-source SOTA levels. Its internal programming benchmarks are 50% better than the previous generation, and it can deliver end-to-end production-grade results. It exhibits emergent cybersecurity capabilities, with the CyberGym vulnerability-discovery benchmark achieving the current best. It supports a 1M-token context window, 128K-token output, and three tiers of deep reasoning, making it suitable for long-range, complex tasks. - Developer: Z.AI - Context window: 1,048,576 tokens - Input modalities: text - Capabilities: Thinking, Tool calling, Structured outputs, Prompt caching - Release date: 2026-10 - Pricing: $2.35/M input tokens, $8.23/M output tokens, $0.59/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"glm-5.3-prime","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"glm-5.3-prime","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/glm-5.3-prime - Try in browser: https://aihubmix.com/playground?model=glm-5.3-prime - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/glm-5.3-prime.6532c100.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/glm-5.3-prime/llms.txt --- # GPT 6.1 Sol Model id on AIHubMix: `gpt-6.1-sol` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-6.1 Sol is OpenAI's latest model in the GPT-6 series, specifically designed for complex programming and agent workflows. GPT-6.1 Sol's performance in complex programming, computer operations, and professional work tasks is close to Astra, but at a lower cost. You can compare the two on your own tasks to weigh quality against cost. - Developer: OpenAI - Context window: 1,050,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Computer use, File search, Structured outputs, Prompt caching, Server-side sessions - Release date: 2026-09-30 - Pricing: $2/M input tokens, $10/M output tokens, $0.1/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-6.1-sol","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-6.1-sol","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-6.1-sol - Try in browser: https://aihubmix.com/playground?model=gpt-6.1-sol - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-6.1-sol.dc9a865a.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-6.1-sol/llms.txt --- # Grok Imagine Image 2.0 Model id on AIHubMix: `grok-imagine-image-2.0` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > grok-imagine-image-2.0 is Grok's text-to-image model, supporting image generation and image editing using up to 5 reference images. - Developer: Grok - Input modalities: text, image - Release date: 2026-10 - Pricing: per-generation — see https://aihubmix.com/model/grok-imagine-image-2.0 ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"grok-imagine-image-2.0","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string|string|null | no | | | | `async` | boolean | no | false | | | `image` | - | no | | | | `images` | array,null | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `prompt` | string | yes | | | | `resolution` | string|null | no | | | | `response_format` | string,null | no | "url" | url, b64_json, null | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/grok-imagine-image-2.0/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/grok-imagine-image-2.0/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/grok-imagine-image-2.0 - Try in browser: https://aihubmix.com/playground?model=grok-imagine-image-2.0 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/grok-imagine-image-2.0/llms.txt --- # Jev 1.14 Model id on AIHubMix: `jev-1.14` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Jev 1.14 is TypeSafe’s structured decision model for classification, scoring, and agent routing, returning typed results with calibrated probabilities. Compared with Jev 1.13, it brings improvements in multilingual support and adds experimental bounding-box outputs for images. - Developer: TypeSafe - Context window: 64,000 tokens - Capabilities: Structured decision - Release date: 2026-10-15 - Pricing: $0.0462/M input tokens **This is a decision model, not a chat model.** It has no `/v1/chat/completions` endpoint — sending `messages` to it returns 404. You send a `state` (the text to judge) plus a map of typed `questions`, and get typed answers back. ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/systemone` — structured decision (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Request `{model, state, questions}`. Each entry in `questions` is named by you and has a `type` of `noul`, `choice` or `score`, an `instructions` string, and — for the latter two — a `criteria` whose **shape differs by type**: - `noul` — no `criteria`. Returns the probability that `instructions` holds. - `choice` — `criteria` is an **object**: option name → what that option means. Returns one option name. - `score` — `criteria` is an **array**: the ordered levels, lowest first. Returns the level index. Passing an array where an object is expected (or vice versa) is an HTTP 400, not a coercion — the two shapes are not interchangeable. ## Example ```bash # Structured decision: POST /v1/systemone — every question is evaluated in parallel and in # isolation against the same state, and comes back typed. No text generation, nothing to parse. # The answers{} map is keyed by your own question names; each answer holds its value under a # key named after its type: # noul -> { type, noul } probability of "yes", 0..1 # choice -> { type, choice, probabilities, confidence } choice is one of your criteria keys # score -> { type, score, legend, probabilities, confidence } legend maps level index -> description # usage carries input_tokens / output_tokens. curl https://aihubmix.com/v1/systemone \ -H "Content-Type: application/json" \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -d '{ "model": "jev-1.14", "state": "Hi, I have been trying to connect my Stripe account for 3 days and it keeps failing. I am losing sales. Please help ASAP.", "questions": { "department": { "type": "choice", "instructions": "Which team should handle this", "criteria": { "billing": "Payment or subscription issues", "technical": "Bugs or integration problems", "sales": "Pricing or account questions" } }, "frustration": { "type": "score", "instructions": "How frustrated the customer appears", "criteria": [ "Calm, just stating facts", "Frustrated but civil", "Very angry, strong language" ] }, "is_urgent": { "type": "noul", "instructions": "The message conveys urgency or time-sensitivity" } } }' ``` ## Response Always a single JSON object — no streaming, no text to parse. `answers` is keyed by **your own** question names, and each answer carries its value under a key named after its type: ```json { "model": "jev-1.13.0", "answers": { "is_urgent": {"type":"noul","noul":0.99}, "department": {"type":"choice","choice":"billing","confidence":0.53,"probabilities":{"billing":0.69,"sales":0.0,"technical":0.31}}, "frustration": {"type":"score","score":1.0,"confidence":1.0,"legend":{"0":"Calm, just stating facts","1":"Frustrated but civil","2":"Very angry, strong language"},"probabilities":{"0":0.0,"1":1.0,"2":0.0}} }, "usage": {"input_tokens":424,"output_tokens":73} } ``` - `noul` — probability in `0..1` that the statement holds. No threshold is baked in; pick your own. - `choice` — one of **your** `criteria` keys, plus the full probability distribution over them. - `score` — the level **index**, `0`-based, counted from your `criteria` array (serialized as a number, so `1` may arrive as `1.0`). `legend` echoes your levels back so the index is readable without holding on to the request. - `confidence` accompanies `choice` and `score` only; `noul` is already a probability. - `model` echoes the resolved upstream build (`jev-latest` resolves to a pinned version here), which is not the id you send. Every question is evaluated in parallel and in isolation against the same `state`, so answers never influence each other and question order carries no meaning. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/jev-1.14 - Try in browser: https://aihubmix.com/playground?model=jev-1.14 - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/jev-1.14.8d058cf3.json — request and response fields with types, enums and official sources. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — `generateDecisionCode()` renders the snippet above in 7 languages, and `buildDecisionBody()` is the same function the Playground calls to build the request it actually sends, so generated snippets and real requests cannot diverge - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/jev-1.14/llms.txt --- # MiniMax M3.1 Flash Model id on AIHubMix: `minimax-m3.1-flash` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > MiniMax-M3.1-Flash is MiniMax's latest flagship model, aimed at office automation, software engineering, visual reasoning, and professional analytics. It combines multi-step task execution with front-end and presentation-creation capabilities, supporting workflows that transform information and visual references into tangible deliverables. - Developer: Minimax - Context window: 1,000,000 tokens - Input modalities: text, image, video - Capabilities: Thinking, Streaming, Tool calling, Web search, Prompt caching - Release date: 2026-10-19 - Pricing: $0.165/M input tokens, $0.66/M output tokens, $0.0165/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"minimax-m3.1-flash","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"minimax-m3.1-flash","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/minimax-m3.1-flash - Try in browser: https://aihubmix.com/playground?model=minimax-m3.1-flash - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/minimax-m3.1-flash.c6ad1a5c.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/minimax-m3.1-flash/llms.txt --- # Gemini 4 Argon Model id on AIHubMix: `gemini-4-argon` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Gemini 4 Argon is Google DeepMind’s multimodal reasoning model for complex software engineering, financial research, legal drafting, and cybersecurity defense. It supports sustained reasoning across long, multi-step workflows, combining coding capabilities with chart and long-video understanding. It can also autonomously discover, validate, and patch software vulnerabilities. - Developer: Google - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio, PDF - Capabilities: Thinking, Streaming, Tool calling, Web search, URL context, Code interpreter, Computer use, File search, Structured outputs, Prompt caching - Pricing: $2/M input tokens, $10/M output tokens, $0.1/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-4-argon","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gemini-4-argon","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gemini-4-argon - Try in browser: https://aihubmix.com/playground?model=gemini-4-argon - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gemini-4-argon.c768758f.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-4-argon/llms.txt --- # Gemini Nano Banana 2.1 Model id on AIHubMix: `gemini-nano-banana-2.1` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Gemini Nano Banana 2.1 (gemini-nano-banana-2.1) is Google's latest high-efficiency image generation and conversational editing model. As an upgraded version of Nano Banana 2 (Gemini 3.1 Flash Image), it significantly improves visual quality, prompt adherence, role consistency in multi-turn conversations, and text rendering while maintaining Flash-level speed and cost efficiency. It can be regarded as a more efficient counterpart to Gemini 3 Pro Image (Nano Banana Pro). - Developer: Google - Input modalities: text, image - Release date: 2026-10-07 - Pricing: per-generation — see https://aihubmix.com/model/gemini-nano-banana-2.1 ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/edits` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-nano-banana-2.1","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string|null | no | null | | | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `image` | - | no | | | | `images` | array,null | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `prompt` | string | yes | | | | `resolution` | string | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/gemini-nano-banana-2.1/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/gemini-nano-banana-2.1/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/gemini-nano-banana-2.1 - Try in browser: https://aihubmix.com/playground?model=gemini-nano-banana-2.1 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-nano-banana-2.1/llms.txt --- # Muse Wonder Model id on AIHubMix: `muse-wonder` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Muse Wonder is Meta’s model for complex coding, long-running agents, and computer use. With a 1M-token context window, it plans multi-step tasks, delegates to sub-agents, and combines Active Memory with Active Focus to retain relevant context and sustain progress on challenging tasks. - Developer: Meta - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio, PDF - Capabilities: Thinking, Streaming, Tool calling, Web search, Computer use, Structured outputs, Citations, Prompt caching, Background mode, Server-side sessions - Release date: 2026-09-03 - Pricing: $4.4/M input tokens, $22/M output tokens, $0.22/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"muse-wonder","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"muse-wonder","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/muse-wonder - Try in browser: https://aihubmix.com/playground?model=muse-wonder - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/muse-wonder.2fbfa656.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/muse-wonder/llms.txt --- # Claude Haiku 5.5 Model id on AIHubMix: `claude-haiku-5-5` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Claude Haiku 5.5 is Anthropic’s next-generation product, designed for high-throughput, cost-sensitive tasks. It can reliably handle fast and repetitive workloads such as summarization, context compression, database queries, and classification requests. It is also well suited to be used as a sub-agent model in programming workflows alongside Opus 5.5 and Sonnet 5.5. - Developer: Anthropic - Context window: 1,000,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Computer use, Memory tool, Document input, PDF input, Structured outputs, Citations, Prompt caching - Release date: 2026-10-08 - Pricing: $0.1/M input tokens, $0.5/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"claude-haiku-5-5","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"claude-haiku-5-5","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/claude-haiku-5-5 - Try in browser: https://aihubmix.com/playground?model=claude-haiku-5-5 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/claude-haiku-5-5.896250b0.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/claude-haiku-5-5/llms.txt --- # Elephant 1 (free) Model id on AIHubMix: `elephant-1-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Elephant 1 is a 9B decision model developed by Base Labs, Baseten's research team, and built on the Qwen3.5-9B architecture. Featuring a context length of 16,384 tokens, this model is designed to analyze a given state in text or JSON format alongside typed questions like yes/no, multiple choice, or ordinal queries. It offers a specialized and efficient solution for automated decision-making workflows. - Developer: Baseten - Pricing: free tier / see model page **This is a decision model, not a chat model.** It has no `/v1/chat/completions` endpoint — sending `messages` to it returns 404. You send a `state` (the text to judge) plus a map of typed `questions`, and get typed answers back. ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/systemone` — structured decision (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Request `{model, state, questions}`. Each entry in `questions` is named by you and has a `type` of `noul`, `choice` or `score`, an `instructions` string, and — for the latter two — a `criteria` whose **shape differs by type**: - `noul` — no `criteria`. Returns the probability that `instructions` holds. - `choice` — `criteria` is an **object**: option name → what that option means. Returns one option name. - `score` — `criteria` is an **array**: the ordered levels, lowest first. Returns the level index. Passing an array where an object is expected (or vice versa) is an HTTP 400, not a coercion — the two shapes are not interchangeable. ## Example ```bash # Structured decision: POST /v1/systemone — every question is evaluated in parallel and in # isolation against the same state, and comes back typed. No text generation, nothing to parse. # The answers{} map is keyed by your own question names; each answer holds its value under a # key named after its type: # noul -> { type, noul } probability of "yes", 0..1 # choice -> { type, choice, probabilities, confidence } choice is one of your criteria keys # score -> { type, score, legend, probabilities, confidence } legend maps level index -> description # usage carries input_tokens / output_tokens. curl https://aihubmix.com/v1/systemone \ -H "Content-Type: application/json" \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -d '{ "model": "elephant-1-free", "state": "Hi, I have been trying to connect my Stripe account for 3 days and it keeps failing. I am losing sales. Please help ASAP.", "questions": { "department": { "type": "choice", "instructions": "Which team should handle this", "criteria": { "billing": "Payment or subscription issues", "technical": "Bugs or integration problems", "sales": "Pricing or account questions" } }, "frustration": { "type": "score", "instructions": "How frustrated the customer appears", "criteria": [ "Calm, just stating facts", "Frustrated but civil", "Very angry, strong language" ] }, "is_urgent": { "type": "noul", "instructions": "The message conveys urgency or time-sensitivity" } } }' ``` ## Response Always a single JSON object — no streaming, no text to parse. `answers` is keyed by **your own** question names, and each answer carries its value under a key named after its type: ```json { "model": "jev-1.13.0", "answers": { "is_urgent": {"type":"noul","noul":0.99}, "department": {"type":"choice","choice":"billing","confidence":0.53,"probabilities":{"billing":0.69,"sales":0.0,"technical":0.31}}, "frustration": {"type":"score","score":1.0,"confidence":1.0,"legend":{"0":"Calm, just stating facts","1":"Frustrated but civil","2":"Very angry, strong language"},"probabilities":{"0":0.0,"1":1.0,"2":0.0}} }, "usage": {"input_tokens":424,"output_tokens":73} } ``` - `noul` — probability in `0..1` that the statement holds. No threshold is baked in; pick your own. - `choice` — one of **your** `criteria` keys, plus the full probability distribution over them. - `score` — the level **index**, `0`-based, counted from your `criteria` array (serialized as a number, so `1` may arrive as `1.0`). `legend` echoes your levels back so the index is readable without holding on to the request. - `confidence` accompanies `choice` and `score` only; `noul` is already a probability. - `model` echoes the resolved upstream build (`jev-latest` resolves to a pinned version here), which is not the id you send. Every question is evaluated in parallel and in isolation against the same `state`, so answers never influence each other and question order carries no meaning. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/elephant-1-free - Try in browser: https://aihubmix.com/playground?model=elephant-1-free - Full parameter schema: follow `https://aihubmix.com/model-data/index.json` — find this id and fetch its `path` (filenames are content-addressed; do not compose them by hand) - Generate runnable code programmatically: npm `@aihubmix/codegen` — `generateDecisionCode()` renders the snippet above in 7 languages, and `buildDecisionBody()` is the same function the Playground calls to build the request it actually sends, so generated snippets and real requests cannot diverge - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/elephant-1-free/llms.txt --- # Mammoth 1 (free) Model id on AIHubMix: `mammoth-1-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Mammoth 1 is a 27B decision model developed by Base Labs, Baseten's research team, and built on Qwen3.5-27B. Featuring a context length of 16,384 tokens, the model is designed to analyze a given state in text or JSON format and answer typed questions such as yes/no, multiple choice, or ordinal queries. It offers a robust and specialized solution for structured decision-making tasks. - Developer: Baseten - Pricing: free tier / see model page **This is a decision model, not a chat model.** It has no `/v1/chat/completions` endpoint — sending `messages` to it returns 404. You send a `state` (the text to judge) plus a map of typed `questions`, and get typed answers back. ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/systemone` — structured decision (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Request `{model, state, questions}`. Each entry in `questions` is named by you and has a `type` of `noul`, `choice` or `score`, an `instructions` string, and — for the latter two — a `criteria` whose **shape differs by type**: - `noul` — no `criteria`. Returns the probability that `instructions` holds. - `choice` — `criteria` is an **object**: option name → what that option means. Returns one option name. - `score` — `criteria` is an **array**: the ordered levels, lowest first. Returns the level index. Passing an array where an object is expected (or vice versa) is an HTTP 400, not a coercion — the two shapes are not interchangeable. ## Example ```bash # Structured decision: POST /v1/systemone — every question is evaluated in parallel and in # isolation against the same state, and comes back typed. No text generation, nothing to parse. # The answers{} map is keyed by your own question names; each answer holds its value under a # key named after its type: # noul -> { type, noul } probability of "yes", 0..1 # choice -> { type, choice, probabilities, confidence } choice is one of your criteria keys # score -> { type, score, legend, probabilities, confidence } legend maps level index -> description # usage carries input_tokens / output_tokens. curl https://aihubmix.com/v1/systemone \ -H "Content-Type: application/json" \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -d '{ "model": "mammoth-1-free", "state": "Hi, I have been trying to connect my Stripe account for 3 days and it keeps failing. I am losing sales. Please help ASAP.", "questions": { "department": { "type": "choice", "instructions": "Which team should handle this", "criteria": { "billing": "Payment or subscription issues", "technical": "Bugs or integration problems", "sales": "Pricing or account questions" } }, "frustration": { "type": "score", "instructions": "How frustrated the customer appears", "criteria": [ "Calm, just stating facts", "Frustrated but civil", "Very angry, strong language" ] }, "is_urgent": { "type": "noul", "instructions": "The message conveys urgency or time-sensitivity" } } }' ``` ## Response Always a single JSON object — no streaming, no text to parse. `answers` is keyed by **your own** question names, and each answer carries its value under a key named after its type: ```json { "model": "jev-1.13.0", "answers": { "is_urgent": {"type":"noul","noul":0.99}, "department": {"type":"choice","choice":"billing","confidence":0.53,"probabilities":{"billing":0.69,"sales":0.0,"technical":0.31}}, "frustration": {"type":"score","score":1.0,"confidence":1.0,"legend":{"0":"Calm, just stating facts","1":"Frustrated but civil","2":"Very angry, strong language"},"probabilities":{"0":0.0,"1":1.0,"2":0.0}} }, "usage": {"input_tokens":424,"output_tokens":73} } ``` - `noul` — probability in `0..1` that the statement holds. No threshold is baked in; pick your own. - `choice` — one of **your** `criteria` keys, plus the full probability distribution over them. - `score` — the level **index**, `0`-based, counted from your `criteria` array (serialized as a number, so `1` may arrive as `1.0`). `legend` echoes your levels back so the index is readable without holding on to the request. - `confidence` accompanies `choice` and `score` only; `noul` is already a probability. - `model` echoes the resolved upstream build (`jev-latest` resolves to a pinned version here), which is not the id you send. Every question is evaluated in parallel and in isolation against the same `state`, so answers never influence each other and question order carries no meaning. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/mammoth-1-free - Try in browser: https://aihubmix.com/playground?model=mammoth-1-free - Full parameter schema: follow `https://aihubmix.com/model-data/index.json` — find this id and fetch its `path` (filenames are content-addressed; do not compose them by hand) - Generate runnable code programmatically: npm `@aihubmix/codegen` — `generateDecisionCode()` renders the snippet above in 7 languages, and `buildDecisionBody()` is the same function the Playground calls to build the request it actually sends, so generated snippets and real requests cannot diverge - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/mammoth-1-free/llms.txt --- # Qwen Image 2.1 Pro Model id on AIHubMix: `qwen-image-2.1-pro` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Qwen-Image-2.1-Pro is a unified text-to-image generation and image-editing model in the Tongyi Qianwen (Qwen) series. Its visual generation component contains only 7 billion parameters (a 32-layer single-stream DiT), achieving a good balance between generation quality, inference efficiency, and versatility. This release includes four key improvements: - Compact and efficient: adopts a lightweight architecture that combines hybrid-granularity attention and prefix KV cache reuse to deliver high-quality image generation with low computational cost. - Native transparency support, unified generation and editing: can generate regular or transparent (RGBA) images from text, edit transparent layers, and extract subjects from photos—all handled by a single model. - Versatile editing: supports up to 10 reference images, allows specifying local edit regions via circular or freehand annotations or separate masks, and preserves identity features of people and products. - Photorealistic textures and refined aesthetics: improved typography, portrait lighting, and detail rendering for more visually appealing results. - Developer: Qwen - Input modalities: text, image - Release date: 2026-10-10 - Pricing: per-generation — see https://aihubmix.com/model/qwen-image-2.1-pro ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/edits` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen-image-2.1-pro","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `image` | - | no | | | | `images` | array,null | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `negative_prompt` | string,null | no | | | | `output_format` | string,null | no | "png" | png, jpeg, webp, null | | `prompt` | string | yes | | | | `response_format` | string,null | no | "url" | url, b64_json, null | | `seed` | integer,null | no | | | | `size` | string|null | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/qwen-image-2.1-pro/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/qwen-image-2.1-pro/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/qwen-image-2.1-pro - Try in browser: https://aihubmix.com/playground?model=qwen-image-2.1-pro - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen-image-2.1-pro/llms.txt --- # Qwen Image 2.1 Turbo Model id on AIHubMix: `qwen-image-2.1-turbo` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Qwen-Image-2.1-Turbo is currently the best-balanced and most cost-effective open-source image generation model in the Qwen-Image series, offering four core advantages: a lightweight model with strong performance and excellent value; native transparency support with integrated generation and editing; flexible editing—supporting multiple reference images, local control, high-fidelity restoration, and comprehensive editing capabilities; and realistic textures, elegant typography, and refined portrait quality. - Developer: Qwen - Input modalities: text, image - Release date: 2026-10-10 - Pricing: per-generation — see https://aihubmix.com/model/qwen-image-2.1-turbo ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/edits` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen-image-2.1-turbo","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `image` | - | no | | | | `images` | array,null | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `negative_prompt` | string,null | no | | | | `output_format` | string,null | no | "png" | png, jpeg, webp, null | | `prompt` | string | yes | | | | `response_format` | string,null | no | "url" | url, b64_json, null | | `seed` | integer,null | no | | | | `size` | string|null | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/qwen-image-2.1-turbo/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/qwen-image-2.1-turbo/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/qwen-image-2.1-turbo - Try in browser: https://aihubmix.com/playground?model=qwen-image-2.1-turbo - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen-image-2.1-turbo/llms.txt --- # Span 01 Lite (free) Model id on AIHubMix: `span-01-lite-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Span-01 Lite is a free, lightweight behavior scoring model developed by Respan. This model analyzes text based on plain-language behaviors that you define and returns the probability of each behavior being present. It offers an efficient and accessible way to evaluate and score content according to your custom criteria. - Developer: Respan - Pricing: free tier / see model page **This is a decision model, not a chat model.** It has no `/v1/chat/completions` endpoint — sending `messages` to it returns 404. You send a `state` (the text to judge) plus a map of typed `questions`, and get typed answers back. ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/systemone` — structured decision (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Request `{model, state, questions}`. Each entry in `questions` is named by you and has a `type` of `noul`, `choice` or `score`, an `instructions` string, and — for the latter two — a `criteria` whose **shape differs by type**: - `noul` — no `criteria`. Returns the probability that `instructions` holds. - `choice` — `criteria` is an **object**: option name → what that option means. Returns one option name. - `score` — `criteria` is an **array**: the ordered levels, lowest first. Returns the level index. Passing an array where an object is expected (or vice versa) is an HTTP 400, not a coercion — the two shapes are not interchangeable. ## Example ```bash # Structured decision: POST /v1/systemone — every question is evaluated in parallel and in # isolation against the same state, and comes back typed. No text generation, nothing to parse. # The answers{} map is keyed by your own question names; each answer holds its value under a # key named after its type: # noul -> { type, noul } probability of "yes", 0..1 # choice -> { type, choice, probabilities, confidence } choice is one of your criteria keys # score -> { type, score, legend, probabilities, confidence } legend maps level index -> description # usage carries input_tokens / output_tokens. curl https://aihubmix.com/v1/systemone \ -H "Content-Type: application/json" \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -d '{ "model": "span-01-lite-free", "state": "Hi, I have been trying to connect my Stripe account for 3 days and it keeps failing. I am losing sales. Please help ASAP.", "questions": { "department": { "type": "choice", "instructions": "Which team should handle this", "criteria": { "billing": "Payment or subscription issues", "technical": "Bugs or integration problems", "sales": "Pricing or account questions" } }, "frustration": { "type": "score", "instructions": "How frustrated the customer appears", "criteria": [ "Calm, just stating facts", "Frustrated but civil", "Very angry, strong language" ] }, "is_urgent": { "type": "noul", "instructions": "The message conveys urgency or time-sensitivity" } } }' ``` ## Response Always a single JSON object — no streaming, no text to parse. `answers` is keyed by **your own** question names, and each answer carries its value under a key named after its type: ```json { "model": "jev-1.13.0", "answers": { "is_urgent": {"type":"noul","noul":0.99}, "department": {"type":"choice","choice":"billing","confidence":0.53,"probabilities":{"billing":0.69,"sales":0.0,"technical":0.31}}, "frustration": {"type":"score","score":1.0,"confidence":1.0,"legend":{"0":"Calm, just stating facts","1":"Frustrated but civil","2":"Very angry, strong language"},"probabilities":{"0":0.0,"1":1.0,"2":0.0}} }, "usage": {"input_tokens":424,"output_tokens":73} } ``` - `noul` — probability in `0..1` that the statement holds. No threshold is baked in; pick your own. - `choice` — one of **your** `criteria` keys, plus the full probability distribution over them. - `score` — the level **index**, `0`-based, counted from your `criteria` array (serialized as a number, so `1` may arrive as `1.0`). `legend` echoes your levels back so the index is readable without holding on to the request. - `confidence` accompanies `choice` and `score` only; `noul` is already a probability. - `model` echoes the resolved upstream build (`jev-latest` resolves to a pinned version here), which is not the id you send. Every question is evaluated in parallel and in isolation against the same `state`, so answers never influence each other and question order carries no meaning. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/span-01-lite-free - Try in browser: https://aihubmix.com/playground?model=span-01-lite-free - Full parameter schema: follow `https://aihubmix.com/model-data/index.json` — find this id and fetch its `path` (filenames are content-addressed; do not compose them by hand) - Generate runnable code programmatically: npm `@aihubmix/codegen` — `generateDecisionCode()` renders the snippet above in 7 languages, and `buildDecisionBody()` is the same function the Playground calls to build the request it actually sends, so generated snippets and real requests cannot diverge - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/span-01-lite-free/llms.txt --- # Claude Opus 5.5 Model id on AIHubMix: `claude-opus-5-5` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Claude Opus 5.5 is Anthropic’s latest flagship model. In Anthropic’s comprehensive alignment testing, Opus 5.5 is the strongest-performing model tested to date. It also includes the safety safeguards Anthropic has developed for its most capable models. - Developer: Anthropic - Context window: 1,000,000 tokens - Input modalities: text, image, PDF - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Computer use, Memory tool, Structured outputs, Citations, Prompt caching - Release date: 2026-09-23 - Pricing: $4/M input tokens, $20/M output tokens, $0.2/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"claude-opus-5-5","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"claude-opus-5-5","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/claude-opus-5-5 - Try in browser: https://aihubmix.com/playground?model=claude-opus-5-5 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/claude-opus-5-5.eb93e5ad.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/claude-opus-5-5/llms.txt --- # Claude Sonnet 5.5 Model id on AIHubMix: `claude-sonnet-5-5` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Claude Sonnet 5.5 is the next-generation product in Anthropic’s Sonnet model series; Claude Sonnet 5.5 offers the best balance between speed and intelligence. The following five major changes will affect code currently running on Claude Sonnet 5: 1. Using "between_tools" can disable up-front thinking. 2. Forcing the use of tools will result in an error. 3. Thinking content blocks are bound to the model and the conversation. 4. On the Claude API and Google Cloud, the older "computer_20251124" computer tool usage is no longer accepted. 5. The advisor tool will refuse to accept Claude Opus 4.8, Claude Opus 4.7, and Claude Sonnet 5 as advisor models. Use between_tools to disable up-front thinking. Forcing tool use will return an error. Thinking blocks are bound to the model and the conversation. On the Claude API and Google Cloud, early computer_20251124 computer tool usage is no longer accepted. The advisor tool no longer accepts Claude Opus 4.8, Claude Opus 4.7, and Claude Sonnet 5 as advisors. - Developer: Anthropic - Context window: 1,000,000 tokens - Input modalities: text, image, PDF - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Computer use, Memory tool, Structured outputs, Citations, Prompt caching - Release date: 2026-09-29 - Pricing: $2/M input tokens, $10/M output tokens, $0.1/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"claude-sonnet-5-5","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"claude-sonnet-5-5","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/claude-sonnet-5-5 - Try in browser: https://aihubmix.com/playground?model=claude-sonnet-5-5 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/claude-sonnet-5-5.2dfbdedf.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/claude-sonnet-5-5/llms.txt --- # GPT 6 Luna Model id on AIHubMix: `gpt-6-luna` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-6 Luna is OpenAI's latest and most efficient model, designed for focused, high-throughput tasks. - Developer: OpenAI - Context window: 1,050,000 tokens - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Computer use, File search, Structured outputs, Structured decision, Prompt caching - Release date: 2026-09-23 - Pricing: $0.1/M input tokens, $0.5/M output tokens, $0.01/M cached input **This is a decision model, not a chat model.** It has no `/v1/chat/completions` endpoint — sending `messages` to it returns 404. You send a `state` (the text to judge) plus a map of typed `questions`, and get typed answers back. ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/systemone` — structured decision (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Request `{model, state, questions}`. Each entry in `questions` is named by you and has a `type` of `noul`, `choice` or `score`, an `instructions` string, and — for the latter two — a `criteria` whose **shape differs by type**: - `noul` — no `criteria`. Returns the probability that `instructions` holds. - `choice` — `criteria` is an **object**: option name → what that option means. Returns one option name. - `score` — `criteria` is an **array**: the ordered levels, lowest first. Returns the level index. Passing an array where an object is expected (or vice versa) is an HTTP 400, not a coercion — the two shapes are not interchangeable. ## Example ```bash # Structured decision: POST /v1/systemone — every question is evaluated in parallel and in # isolation against the same state, and comes back typed. No text generation, nothing to parse. # The answers{} map is keyed by your own question names; each answer holds its value under a # key named after its type: # noul -> { type, noul } probability of "yes", 0..1 # choice -> { type, choice, probabilities, confidence } choice is one of your criteria keys # score -> { type, score, legend, probabilities, confidence } legend maps level index -> description # usage carries input_tokens / output_tokens. curl https://aihubmix.com/v1/systemone \ -H "Content-Type: application/json" \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -d '{ "model": "gpt-6-luna", "state": "Hi, I have been trying to connect my Stripe account for 3 days and it keeps failing. I am losing sales. Please help ASAP.", "questions": { "department": { "type": "choice", "instructions": "Which team should handle this", "criteria": { "billing": "Payment or subscription issues", "technical": "Bugs or integration problems", "sales": "Pricing or account questions" } }, "frustration": { "type": "score", "instructions": "How frustrated the customer appears", "criteria": [ "Calm, just stating facts", "Frustrated but civil", "Very angry, strong language" ] }, "is_urgent": { "type": "noul", "instructions": "The message conveys urgency or time-sensitivity" } } }' ``` ## Response Always a single JSON object — no streaming, no text to parse. `answers` is keyed by **your own** question names, and each answer carries its value under a key named after its type: ```json { "model": "jev-1.13.0", "answers": { "is_urgent": {"type":"noul","noul":0.99}, "department": {"type":"choice","choice":"billing","confidence":0.53,"probabilities":{"billing":0.69,"sales":0.0,"technical":0.31}}, "frustration": {"type":"score","score":1.0,"confidence":1.0,"legend":{"0":"Calm, just stating facts","1":"Frustrated but civil","2":"Very angry, strong language"},"probabilities":{"0":0.0,"1":1.0,"2":0.0}} }, "usage": {"input_tokens":424,"output_tokens":73} } ``` - `noul` — probability in `0..1` that the statement holds. No threshold is baked in; pick your own. - `choice` — one of **your** `criteria` keys, plus the full probability distribution over them. - `score` — the level **index**, `0`-based, counted from your `criteria` array (serialized as a number, so `1` may arrive as `1.0`). `legend` echoes your levels back so the index is readable without holding on to the request. - `confidence` accompanies `choice` and `score` only; `noul` is already a probability. - `model` echoes the resolved upstream build (`jev-latest` resolves to a pinned version here), which is not the id you send. Every question is evaluated in parallel and in isolation against the same `state`, so answers never influence each other and question order carries no meaning. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-6-luna - Try in browser: https://aihubmix.com/playground?model=gpt-6-luna - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-6-luna.03ea4c4d.json — request and response fields with types, enums and official sources. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — `generateDecisionCode()` renders the snippet above in 7 languages, and `buildDecisionBody()` is the same function the Playground calls to build the request it actually sends, so generated snippets and real requests cannot diverge - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-6-luna/llms.txt --- # GPT 6 Sol Model id on AIHubMix: `gpt-6-sol` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-6 Sol is OpenAI's latest model in the GPT-6 series, specifically designed for complex programming and agent workflows. - Developer: OpenAI - Context window: 1,050,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Computer use, File search, Structured outputs, Prompt caching - Release date: 2026-09-23 - Pricing: $2/M input tokens, $10/M output tokens, $0.2/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-6-sol","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-6-sol","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-6-sol - Try in browser: https://aihubmix.com/playground?model=gpt-6-sol - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-6-sol.095ca7f2.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-6-sol/llms.txt --- # Mistral Large 4.0 Model id on AIHubMix: `mistral-large-4-0` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Mistral Large 4 is Mistral AI’s frontier multimodal model with text and image input, designed for complex reasoning, coding, and agentic workflows. It supports tool calling and structured outputs, with a 512K-token context window and up to 256K output tokens, making it well suited for long-context and multi-step tasks. Compared with earlier Mistral large models, Mistral Large 4 places greater emphasis on multimodal understanding, extended context, and agentic execution. - Developer: Mistral - Context window: 1,000,000 tokens - Input modalities: text, image - Capabilities: Thinking, Tool calling, Document input, Structured outputs - Release date: 2026-10-06 - Pricing: $1.36/M input tokens, $4.18/M output tokens, $0.1399/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"mistral-large-4-0","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"mistral-large-4-0","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/mistral-large-4-0 - Try in browser: https://aihubmix.com/playground?model=mistral-large-4-0 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/mistral-large-4-0.5c5e45cc.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/mistral-large-4-0/llms.txt --- # Gemini 3.8 Flash Lite Tts Model id on AIHubMix: `gemini-3.8-flash-lite-tts` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Gemini 3.8 Flash-Lite TTS (gemini-3.8-flash-lite-tts) is Google's fast, cost-efficient workhorse text-to-speech model, built to replace gemini-3.1-flash-tts-preview for high-throughput production workloads. - Developer: Google - Context window: 8,192 tokens - Input modalities: text - Release date: 2026-09-24 - Pricing: $0.5/M input tokens, $6/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-3.8-flash-lite-tts","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gemini-3.8-flash-lite-tts","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gemini-3.8-flash-lite-tts - Try in browser: https://aihubmix.com/playground?model=gemini-3.8-flash-lite-tts - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gemini-3.8-flash-lite-tts.dbe90757.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-3.8-flash-lite-tts/llms.txt --- # Gemini 3.8 Flash Tts Model id on AIHubMix: `gemini-3.8-flash-tts` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Gemini 3.8 Flash TTS (gemini-3.8-flash-tts) is Google's 3.8 Flash text-to-speech audio model, designed to deliver studio-grade sound fidelity, expressive performance, and long-form stability. - Developer: Google - Context window: 8,192 tokens - Input modalities: text - Release date: 2026-09-24 - Pricing: $0.5/M input tokens, $9/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-3.8-flash-tts","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gemini-3.8-flash-tts","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gemini-3.8-flash-tts - Try in browser: https://aihubmix.com/playground?model=gemini-3.8-flash-tts - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gemini-3.8-flash-tts.03a9b7df.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-3.8-flash-tts/llms.txt --- # Grok 4.7 Model id on AIHubMix: `grok-4.7` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Grok 4.7 is xAI’s flagship model for coding, agentic tasks, and professional knowledge work. Compared with Grok 4.6, it uses a larger base model and longer reinforcement-learning training, improving long-horizon task execution, self-verification, context management, and document and presentation creation. - Developer: Grok - Context window: 500,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Document input, Structured outputs, Prompt caching, Server-side sessions - Release date: 2026-09-22 - Pricing: $2/M input tokens, $6/M output tokens, $0.5/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"grok-4.7","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"grok-4.7","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/grok-4.7 - Try in browser: https://aihubmix.com/playground?model=grok-4.7 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/grok-4.7.c53e186d.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/grok-4.7/llms.txt --- # MiMo V2.6 Flash Model id on AIHubMix: `mimo-v2.6-flash` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > mimo-v2.6-flash is Xiaomi's latest model series. A fully multimodal, highly intelligent, low-cost efficient inference model, it is the best-balanced choice for high-frequency invocations and large-scale tasks in professional office settings. - Developer: Xiaomi - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Release date: 2026-09-22 - Pricing: $0.155/M input tokens, $0.31/M output tokens, $0.0031/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"mimo-v2.6-flash","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"mimo-v2.6-flash","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/mimo-v2.6-flash - Try in browser: https://aihubmix.com/playground?model=mimo-v2.6-flash - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/mimo-v2.6-flash.b04d1f2d.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/mimo-v2.6-flash/llms.txt --- # MiMo V2.6 Pro Model id on AIHubMix: `mimo-v2.6-pro` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > MiMo-V2.6-Pro is Xiaomi’s trillion-parameter, natively omni-modal flagship reasoning model for complex projects, long-horizon agentic tasks, coding, research, and professional knowledge work. It supports a 1M-token context window and up to 128K output tokens. Compared with V2.5, it delivers stronger reasoning and long-task performance while retaining the same API pricing. - Developer: Xiaomi - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Release date: 2026-09-22 - Pricing: $0.48/M input tokens, $0.96/M output tokens, $0.0038/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"mimo-v2.6-pro","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"mimo-v2.6-pro","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/mimo-v2.6-pro - Try in browser: https://aihubmix.com/playground?model=mimo-v2.6-pro - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/mimo-v2.6-pro.81aeaf60.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/mimo-v2.6-pro/llms.txt --- # MiMo V2.6 Pro Ultraspeed Model id on AIHubMix: `mimo-v2.6-pro-ultraspeed` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > MiMo-V2.6-Pro-UltraSpeed combines the flagship performance of MiMo-V2.6-Pro and offers up to 20× output speed to serve highly real-time interactive and latency-sensitive production scenarios. - Developer: Xiaomi - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Release date: 2026-09-22 - Pricing: $4.8/M input tokens, $9.6/M output tokens, $0.0384/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"mimo-v2.6-pro-ultraspeed","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"mimo-v2.6-pro-ultraspeed","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/mimo-v2.6-pro-ultraspeed - Try in browser: https://aihubmix.com/playground?model=mimo-v2.6-pro-ultraspeed - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/mimo-v2.6-pro-ultraspeed.b3c50d02.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/mimo-v2.6-pro-ultraspeed/llms.txt --- # Qwen3.8 Omni Flash Model id on AIHubMix: `qwen3.8-omni-flash` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Qwen3.8 Omni Flash is Alibaba Cloud Qwen's next-generation native multimodal model, supporting 1M-length sequences and able to directly accept text, images, audio, and video as input. It is based on the Qwen3.8-Flash-Next architecture. The model is designed for agent capabilities in real productivity scenarios: while offering agentic abilities such as programming, text knowledge work, and GUI operation, it also achieves significant results in audio/video-centered agentic applications—such as video editing, music video creation, film production and narration, audio/video-to-text-and-image summarization, and audio/video dialogue—that require integrated processing of text, images, audio, and video. It supports 2-channel and 4-channel spatial audio parsing, is compatible with DashScope and OpenAI protocols, and it is recommended to install the accompanying Qwen-MM-Plugins to facilitate agent frameworks' access to native multimodal capabilities. - Developer: Qwen - Context window: 1,000,000 tokens - Input modalities: text, image, audio, video - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Release date: 2026-08-26 - Pricing: $0.1126/M input tokens, $0.38/M output tokens, $0.0141/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen3.8-omni-flash","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"qwen3.8-omni-flash","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/qwen3.8-omni-flash - Try in browser: https://aihubmix.com/playground?model=qwen3.8-omni-flash - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/qwen3.8-omni-flash.452b1d36.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen3.8-omni-flash/llms.txt --- # Decision Model Preview Model id on AIHubMix: `decision-model-preview` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Alibaba Cloud has launched the decision model decision-model-preview. This structured decision model is designed for high-frequency business judgments; it can concurrently perform classification, binary decisions, and scoring based on text or business state, and returns probability distributions and confidence levels. It is suitable for scenarios such as ticket routing, content moderation, agent routing, and result verification. Free for a limited time. - Developer: Qwen - Context window: 64,000 tokens - Release date: 2026-09-24 - Pricing: free tier / see model page **This is a decision model, not a chat model.** It has no `/v1/chat/completions` endpoint — sending `messages` to it returns 404. You send a `state` (the text to judge) plus a map of typed `questions`, and get typed answers back. ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/systemone` — structured decision (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Request `{model, state, questions}`. Each entry in `questions` is named by you and has a `type` of `noul`, `choice` or `score`, an `instructions` string, and — for the latter two — a `criteria` whose **shape differs by type**: - `noul` — no `criteria`. Returns the probability that `instructions` holds. - `choice` — `criteria` is an **object**: option name → what that option means. Returns one option name. - `score` — `criteria` is an **array**: the ordered levels, lowest first. Returns the level index. Passing an array where an object is expected (or vice versa) is an HTTP 400, not a coercion — the two shapes are not interchangeable. ## Example ```bash # Structured decision: POST /v1/systemone — every question is evaluated in parallel and in # isolation against the same state, and comes back typed. No text generation, nothing to parse. # The answers{} map is keyed by your own question names; each answer holds its value under a # key named after its type: # noul -> { type, noul } probability of "yes", 0..1 # choice -> { type, choice, probabilities, confidence } choice is one of your criteria keys # score -> { type, score, legend, probabilities, confidence } legend maps level index -> description # usage carries input_tokens / output_tokens. curl https://aihubmix.com/v1/systemone \ -H "Content-Type: application/json" \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -d '{ "model": "decision-model-preview", "state": "Hi, I have been trying to connect my Stripe account for 3 days and it keeps failing. I am losing sales. Please help ASAP.", "questions": { "department": { "type": "choice", "instructions": "Which team should handle this", "criteria": { "billing": "Payment or subscription issues", "technical": "Bugs or integration problems", "sales": "Pricing or account questions" } }, "frustration": { "type": "score", "instructions": "How frustrated the customer appears", "criteria": [ "Calm, just stating facts", "Frustrated but civil", "Very angry, strong language" ] }, "is_urgent": { "type": "noul", "instructions": "The message conveys urgency or time-sensitivity" } } }' ``` ## Response Always a single JSON object — no streaming, no text to parse. `answers` is keyed by **your own** question names, and each answer carries its value under a key named after its type: ```json { "model": "jev-1.13.0", "answers": { "is_urgent": {"type":"noul","noul":0.99}, "department": {"type":"choice","choice":"billing","confidence":0.53,"probabilities":{"billing":0.69,"sales":0.0,"technical":0.31}}, "frustration": {"type":"score","score":1.0,"confidence":1.0,"legend":{"0":"Calm, just stating facts","1":"Frustrated but civil","2":"Very angry, strong language"},"probabilities":{"0":0.0,"1":1.0,"2":0.0}} }, "usage": {"input_tokens":424,"output_tokens":73} } ``` - `noul` — probability in `0..1` that the statement holds. No threshold is baked in; pick your own. - `choice` — one of **your** `criteria` keys, plus the full probability distribution over them. - `score` — the level **index**, `0`-based, counted from your `criteria` array (serialized as a number, so `1` may arrive as `1.0`). `legend` echoes your levels back so the index is readable without holding on to the request. - `confidence` accompanies `choice` and `score` only; `noul` is already a probability. - `model` echoes the resolved upstream build (`jev-latest` resolves to a pinned version here), which is not the id you send. Every question is evaluated in parallel and in isolation against the same `state`, so answers never influence each other and question order carries no meaning. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/decision-model-preview - Try in browser: https://aihubmix.com/playground?model=decision-model-preview - Full parameter schema: follow `https://aihubmix.com/model-data/index.json` — find this id and fetch its `path` (filenames are content-addressed; do not compose them by hand) - Generate runnable code programmatically: npm `@aihubmix/codegen` — `generateDecisionCode()` renders the snippet above in 7 languages, and `buildDecisionBody()` is the same function the Playground calls to build the request it actually sends, so generated snippets and real requests cannot diverge - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/decision-model-preview/llms.txt --- # Jev 1.13 Model id on AIHubMix: `jev-1.13` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Jev is a structured decision model launched by TypeSafe and is also their first "System One" model. System One models can make structured decisions for software quickly, returning typed selection results rather than freely generated text. They are suitable for routing, classification, and other decision nodes in applications—where fast, predictable answers are more important than generating natural language content. - Developer: TypeSafe - Context window: 64,000 tokens - Capabilities: Structured decision - Release date: 2026-09-17 - Pricing: $0.0462/M input tokens **This is a decision model, not a chat model.** It has no `/v1/chat/completions` endpoint — sending `messages` to it returns 404. You send a `state` (the text to judge) plus a map of typed `questions`, and get typed answers back. ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/systemone` — structured decision (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Request `{model, state, questions}`. Each entry in `questions` is named by you and has a `type` of `noul`, `choice` or `score`, an `instructions` string, and — for the latter two — a `criteria` whose **shape differs by type**: - `noul` — no `criteria`. Returns the probability that `instructions` holds. - `choice` — `criteria` is an **object**: option name → what that option means. Returns one option name. - `score` — `criteria` is an **array**: the ordered levels, lowest first. Returns the level index. Passing an array where an object is expected (or vice versa) is an HTTP 400, not a coercion — the two shapes are not interchangeable. ## Example ```bash # Structured decision: POST /v1/systemone — every question is evaluated in parallel and in # isolation against the same state, and comes back typed. No text generation, nothing to parse. # The answers{} map is keyed by your own question names; each answer holds its value under a # key named after its type: # noul -> { type, noul } probability of "yes", 0..1 # choice -> { type, choice, probabilities, confidence } choice is one of your criteria keys # score -> { type, score, legend, probabilities, confidence } legend maps level index -> description # usage carries input_tokens / output_tokens. curl https://aihubmix.com/v1/systemone \ -H "Content-Type: application/json" \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -d '{ "model": "jev-1.13", "state": "Hi, I have been trying to connect my Stripe account for 3 days and it keeps failing. I am losing sales. Please help ASAP.", "questions": { "department": { "type": "choice", "instructions": "Which team should handle this", "criteria": { "billing": "Payment or subscription issues", "technical": "Bugs or integration problems", "sales": "Pricing or account questions" } }, "frustration": { "type": "score", "instructions": "How frustrated the customer appears", "criteria": [ "Calm, just stating facts", "Frustrated but civil", "Very angry, strong language" ] }, "is_urgent": { "type": "noul", "instructions": "The message conveys urgency or time-sensitivity" } } }' ``` ## Response Always a single JSON object — no streaming, no text to parse. `answers` is keyed by **your own** question names, and each answer carries its value under a key named after its type: ```json { "model": "jev-1.13.0", "answers": { "is_urgent": {"type":"noul","noul":0.99}, "department": {"type":"choice","choice":"billing","confidence":0.53,"probabilities":{"billing":0.69,"sales":0.0,"technical":0.31}}, "frustration": {"type":"score","score":1.0,"confidence":1.0,"legend":{"0":"Calm, just stating facts","1":"Frustrated but civil","2":"Very angry, strong language"},"probabilities":{"0":0.0,"1":1.0,"2":0.0}} }, "usage": {"input_tokens":424,"output_tokens":73} } ``` - `noul` — probability in `0..1` that the statement holds. No threshold is baked in; pick your own. - `choice` — one of **your** `criteria` keys, plus the full probability distribution over them. - `score` — the level **index**, `0`-based, counted from your `criteria` array (serialized as a number, so `1` may arrive as `1.0`). `legend` echoes your levels back so the index is readable without holding on to the request. - `confidence` accompanies `choice` and `score` only; `noul` is already a probability. - `model` echoes the resolved upstream build (`jev-latest` resolves to a pinned version here), which is not the id you send. Every question is evaluated in parallel and in isolation against the same `state`, so answers never influence each other and question order carries no meaning. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/jev-1.13 - Try in browser: https://aihubmix.com/playground?model=jev-1.13 - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/jev-1.13.5d06f8ee.json — request and response fields with types, enums and official sources. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — `generateDecisionCode()` renders the snippet above in 7 languages, and `buildDecisionBody()` is the same function the Playground calls to build the request it actually sends, so generated snippets and real requests cannot diverge - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/jev-1.13/llms.txt --- # Jev Model id on AIHubMix: `jev-latest` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Jev is a structured decision model launched by TypeSafe and is also their first "System One" model. System One models can make structured decisions for software quickly, returning typed selection results rather than freely generated text. They are suitable for routing, classification, and other decision nodes in applications—where fast, predictable answers are more important than generating natural language content. - Developer: TypeSafe - Context window: 64,000 tokens - Capabilities: Structured decision - Release date: 2026-09-17 - Pricing: $0.0462/M input tokens **This is a decision model, not a chat model.** It has no `/v1/chat/completions` endpoint — sending `messages` to it returns 404. You send a `state` (the text to judge) plus a map of typed `questions`, and get typed answers back. ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/systemone` — structured decision (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Request `{model, state, questions}`. Each entry in `questions` is named by you and has a `type` of `noul`, `choice` or `score`, an `instructions` string, and — for the latter two — a `criteria` whose **shape differs by type**: - `noul` — no `criteria`. Returns the probability that `instructions` holds. - `choice` — `criteria` is an **object**: option name → what that option means. Returns one option name. - `score` — `criteria` is an **array**: the ordered levels, lowest first. Returns the level index. Passing an array where an object is expected (or vice versa) is an HTTP 400, not a coercion — the two shapes are not interchangeable. ## Example ```bash # Structured decision: POST /v1/systemone — every question is evaluated in parallel and in # isolation against the same state, and comes back typed. No text generation, nothing to parse. # The answers{} map is keyed by your own question names; each answer holds its value under a # key named after its type: # noul -> { type, noul } probability of "yes", 0..1 # choice -> { type, choice, probabilities, confidence } choice is one of your criteria keys # score -> { type, score, legend, probabilities, confidence } legend maps level index -> description # usage carries input_tokens / output_tokens. curl https://aihubmix.com/v1/systemone \ -H "Content-Type: application/json" \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -d '{ "model": "jev-latest", "state": "Hi, I have been trying to connect my Stripe account for 3 days and it keeps failing. I am losing sales. Please help ASAP.", "questions": { "department": { "type": "choice", "instructions": "Which team should handle this", "criteria": { "billing": "Payment or subscription issues", "technical": "Bugs or integration problems", "sales": "Pricing or account questions" } }, "frustration": { "type": "score", "instructions": "How frustrated the customer appears", "criteria": [ "Calm, just stating facts", "Frustrated but civil", "Very angry, strong language" ] }, "is_urgent": { "type": "noul", "instructions": "The message conveys urgency or time-sensitivity" } } }' ``` ## Response Always a single JSON object — no streaming, no text to parse. `answers` is keyed by **your own** question names, and each answer carries its value under a key named after its type: ```json { "model": "jev-1.13.0", "answers": { "is_urgent": {"type":"noul","noul":0.99}, "department": {"type":"choice","choice":"billing","confidence":0.53,"probabilities":{"billing":0.69,"sales":0.0,"technical":0.31}}, "frustration": {"type":"score","score":1.0,"confidence":1.0,"legend":{"0":"Calm, just stating facts","1":"Frustrated but civil","2":"Very angry, strong language"},"probabilities":{"0":0.0,"1":1.0,"2":0.0}} }, "usage": {"input_tokens":424,"output_tokens":73} } ``` - `noul` — probability in `0..1` that the statement holds. No threshold is baked in; pick your own. - `choice` — one of **your** `criteria` keys, plus the full probability distribution over them. - `score` — the level **index**, `0`-based, counted from your `criteria` array (serialized as a number, so `1` may arrive as `1.0`). `legend` echoes your levels back so the index is readable without holding on to the request. - `confidence` accompanies `choice` and `score` only; `noul` is already a probability. - `model` echoes the resolved upstream build (`jev-latest` resolves to a pinned version here), which is not the id you send. Every question is evaluated in parallel and in isolation against the same `state`, so answers never influence each other and question order carries no meaning. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/jev-latest - Try in browser: https://aihubmix.com/playground?model=jev-latest - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/jev-latest.514f58ac.json — request and response fields with types, enums and official sources. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — `generateDecisionCode()` renders the snippet above in 7 languages, and `buildDecisionBody()` is the same function the Playground calls to build the request it actually sends, so generated snippets and real requests cannot diverge - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/jev-latest/llms.txt --- # Step 5 Preview Model id on AIHubMix: `step-5-preview` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Step 5 Preview is StepFun's next-generation flagship base model designed for real-world tasks, with a focus on programming and professional expertise. - Developer: StepFun - Context window: 1,000,000 tokens - Input modalities: text, image, video - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2026-09-20 - Pricing: $1.08/M input tokens, $3.0888/M output tokens, $0.054/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"step-5-preview","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"step-5-preview","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/step-5-preview - Try in browser: https://aihubmix.com/playground?model=step-5-preview - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/step-5-preview.73c7b48a.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/step-5-preview/llms.txt --- # Doubao Seedream 5.0 Flash Model id on AIHubMix: `doubao-seedream-5-0-flash` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Doubao-Seedream-5.0-flash is a fast image-creation model released by ByteDance that balances speed and cost efficiency; it supports precise editing, layer separation, and transparent-background output, making it suitable for high-frequency, large-scale image production and rapid interactive scenarios. - Developer: ByteDance - Input modalities: text, image - Pricing: per-generation — see https://aihubmix.com/model/doubao-seedream-5-0-flash ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/edits` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"doubao-seedream-5-0-flash","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string,null | no | | 16:9, 9:16, 1:1, 4:3, 3:4, 3:2, … | | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `image` | - | no | | | | `images` | array,null | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `output_format` | string,null | no | "jpeg" | png, jpeg | | `prompt` | string | yes | | | | `response_format` | string,null | no | "url" | url, b64_json, null | | `seed` | integer,null | no | | | | `size` | string,null | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/doubao-seedream-5-0-flash/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/doubao-seedream-5-0-flash/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/doubao-seedream-5-0-flash - Try in browser: https://aihubmix.com/playground?model=doubao-seedream-5-0-flash - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/doubao-seedream-5-0-flash/llms.txt --- # GPT 6 Astra Model id on AIHubMix: `gpt-6-astra` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-6 Astra is OpenAI's newest and most intelligent model, with industry-leading performance in computer operations, web browsing, software engineering, scientific research, and professional work. It excels at executing multi-step workflows across code, browsers, and various professional software. Astra can achieve better results with significantly fewer output tokens, making its estimated API cost per task lower. - Developer: OpenAI - Context window: 1,050,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Computer use, File search, Structured outputs, Prompt caching, Server-side sessions - Release date: 2026-09-05 - Pricing: $10/M input tokens, $50/M output tokens, $1/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/responses` — OpenAI Responses (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-6-astra","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-6-astra","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-6-astra - Try in browser: https://aihubmix.com/playground?model=gpt-6-astra - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-6-astra.3185a7ac.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-6-astra/llms.txt --- # Gemini 3.8 Flash Model id on AIHubMix: `gemini-3.8-flash` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Gemini 3.8 Flash is Google's most intelligent Flash-series model, designed for long-running software engineering tasks, autonomous agents, and complex enterprise workflows, while retaining the Flash series' fast responsiveness and cost-effectiveness. - Developer: Google - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio, PDF - Capabilities: Thinking, Streaming, Tool calling, Web search, URL context, Code interpreter, Computer use, File search, Structured outputs, Prompt caching - Release date: 2026-09-02 - Pricing: $0.75/M input tokens, $3.75/M output tokens, $0.075/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /gemini/v1beta/models/gemini-3.8-flash:generateContent` — Google Gemini (`x-goog-api-key: $AIHUBMIX_API_KEY`; streaming: `:streamGenerateContent?alt=sse`) - `POST /v1/messages` — Anthropic Messages (`x-api-key: $AIHUBMIX_API_KEY` + `anthropic-version: 2023-06-01`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-3.8-flash","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gemini-3.8-flash","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gemini-3.8-flash - Try in browser: https://aihubmix.com/playground?model=gemini-3.8-flash - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gemini-3.8-flash.1b712b10.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-3.8-flash/llms.txt --- # Claude Fable 5.1 Model id on AIHubMix: `claude-fable-5-1` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Claude-fable-5-1 is Anthropic's most capable newly released model, suitable for the most demanding reasoning and long-running agent work. Claude Fable 5.1, while keeping input and output prices the same as Claude Fable 5, reduces cache read prices to one quarter of the original and is more capable for long-running agent programming, multi-step research, and document, spreadsheet, and slide processing. (The model is extremely expensive and not recommended for casual use.) - Developer: Anthropic - Context window: 1,000,000 tokens - Input modalities: text, image, PDF - Capabilities: Thinking, Streaming, Tool calling, Code interpreter, Computer use, Memory tool, Structured outputs, Citations, Prompt caching - Release date: 2026-09-02 - Pricing: $11/M input tokens, $55/M output tokens, $0.275/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/messages` — Anthropic Messages (`x-api-key: $AIHUBMIX_API_KEY` + `anthropic-version: 2023-06-01`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"claude-fable-5-1","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"claude-fable-5-1","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/claude-fable-5-1 - Try in browser: https://aihubmix.com/playground?model=claude-fable-5-1 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/claude-fable-5-1.b8da0872.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/claude-fable-5-1/llms.txt --- # DeepSeek V4.1 Flash Model id on AIHubMix: `deepseek-v4.1-flash` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > DeepSeek-V4.1-Flash model official release. This is the smallest model in DeepSeek’s new model-architecture series, featuring native multimodal visual understanding capabilities. The new architecture is designed to deliver a higher capability ceiling, faster inference speeds, greater throughput, and scalability to larger-parameter models. - Developer: DeepSeek - Context window: 1,000,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2026-09-08 - Pricing: $0.1408/M input tokens, $0.5632/M output tokens, $0.0028/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/messages` — Anthropic Messages (`x-api-key: $AIHUBMIX_API_KEY` + `anthropic-version: 2023-06-01`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"deepseek-v4.1-flash","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"deepseek-v4.1-flash","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/deepseek-v4.1-flash - Try in browser: https://aihubmix.com/playground?model=deepseek-v4.1-flash - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/deepseek-v4.1-flash.449ab7bc.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/deepseek-v4.1-flash/llms.txt --- # GLM 5.3 Flash Model id on AIHubMix: `glm-5.3-flash` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GLM-5.3-Flash is a high-efficiency multimodal model from Z.AI. It supports a context window of roughly 1 million tokens, along with text, image, and video inputs, and includes tool-calling capabilities. It is primarily designed for coding agents, complex reasoning, and long-horizon software engineering tasks. Built on the existing GLM technology stack, the model has been further post-trained and optimized to deliver strong performance while placing greater emphasis on inference efficiency, responsiveness, and cost.The model is offered at a limited-time 50% discount; users are welcome to try it. - Developer: Z.AI - Context window: 1,048,576 tokens - Input modalities: text, image, video - Capabilities: Thinking, Streaming, Tool calling, Prompt caching - Release date: 2026-08-26 - Pricing: $0.1127/M input tokens, $0.3944/M output tokens, $0.0324/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"glm-5.3-flash","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"glm-5.3-flash","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/glm-5.3-flash - Try in browser: https://aihubmix.com/playground?model=glm-5.3-flash - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/glm-5.3-flash.3f7be1b3.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/glm-5.3-flash/llms.txt --- # GPT Live Transcribe Model id on AIHubMix: `gpt-live-transcribe` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > OpenAI's latest realtime speech-to-text model, built for low-latency use — it streams incremental transcripts as audio arrives, ideal for live captions, dictation, and voice interfaces. Supports automatic transcription across 57 languages, with keyword/context hints and a tunable latency-vs-accuracy tradeoff. - Developer: OpenAI - Session kind: realtime WebSocket — speech-to-text transcription - Input modalities: audio, text - Release date: 2026-07-28 - Pricing: see https://aihubmix.com/model/gpt-live-transcribe (realtime is billed on audio + text tokens) ## Endpoints (base URL: https://aihubmix.com) - `GET wss://aihubmix.com/v1/realtime?intent=transcription&model=gpt-live-transcribe` — realtime speech-to-text transcription over WebSocket. Authenticate the handshake with `Authorization: Bearer $AIHUBMIX_API_KEY`. The model id is carried in the handshake query above (with `intent=transcription`); audio is PCM16 / 24kHz / mono in both directions. ## Example ```python # pip install websockets import asyncio import base64 import json import os import websockets # Realtime transcription is a WebSocket session: model must be in the handshake URL. URL = "wss://aihubmix.com/v1/realtime?intent=transcription&model=gpt-live-transcribe" async def main(): # websockets >= 13 uses additional_headers; older versions use extra_headers async with websockets.connect( URL, additional_headers={"Authorization": "Bearer " + os.environ["AIHUBMIX_API_KEY"]} ) as ws: # 1) Configure the transcription session (first frame; server VAD auto-segments speech as you stream) await ws.send(json.dumps({ "type": "session.update", "session": { "type": "transcription", "audio": { "input": { "format": { "type": "audio/pcm", "rate": 24000 }, "transcription": { "model": "gpt-live-transcribe" }, "turn_detection": { "type": "server_vad" } } } } })) flushed = asyncio.Event() # set once the whole stream is sent + flushed # 2) Stream raw PCM16 / 24kHz / mono audio in ~100ms chunks async def send_audio(): with open("audio_pcm16_24k.raw", "rb") as f: pcm = f.read() chunk = 24000 * 2 * 100 // 1000 # 100ms of 16-bit mono samples for i in range(0, len(pcm), chunk): await ws.send(json.dumps({ "type": "input_audio_buffer.append", "audio": base64.b64encode(pcm[i:i + chunk]).decode(), })) await asyncio.sleep(0.1) # simulate realtime pacing # Server VAD segments on pauses; flush any trailing audio at end of stream await ws.send(json.dumps({"type": "input_audio_buffer.commit"})) flushed.set() asyncio.create_task(send_audio()) # 3) Receive events: deltas stream live; each pause finalizes a segment (server VAD) async for msg in ws: evt = json.loads(msg) etype = evt.get("type", "") if etype.endswith("transcription.delta"): print(evt.get("delta", ""), end="", flush=True) elif etype.endswith("transcription.completed"): print("\n[segment]", evt.get("transcript", "")) if flushed.is_set(): # last segment after the final flush → done break elif etype == "error": print("\n[error]", evt.get("error")) break asyncio.run(main()) ``` ## Response Realtime transcription is an **event stream over one WebSocket**, not a single JSON body. After the first `session.update`, stream raw PCM16 / 24kHz / mono audio as `input_audio_buffer.append` frames; server VAD segments speech on pauses. Read events until done: - `*.transcription.delta` — partial text as it is recognized (`delta` field) - `*.transcription.completed` — a finalized segment (`transcript` field) - `error` — a session-level error; the socket closes No HTTP response body and no `stream: true` flag — the transport itself is the stream. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-live-transcribe - Try in browser: https://aihubmix.com/playground?model=gpt-live-transcribe - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-live-transcribe.96833760.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (this realtime example was produced by its `generateRealtimeCode`; 7 languages, media and realtime included); the session frames it builds are the exact frames the gateway receives, so generated snippets and real requests cannot diverge - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-live-transcribe/llms.txt --- # Xiaomi Mimo V2.6 Pro (free) Model id on AIHubMix: `xiaomi-mimo-v2.6-pro-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > xiaomi-mimo-v2.6-pro-free is the open free version of xiaomi-mimo-v2.6-pro. To maintain reliable service, each account is limited to 5 requests per minute, 100 requests per day, and 1 million tokens per day. - Developer: Xiaomi - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Release date: 2026-09-22 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"xiaomi-mimo-v2.6-pro-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"xiaomi-mimo-v2.6-pro-free","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/xiaomi-mimo-v2.6-pro-free - Try in browser: https://aihubmix.com/playground?model=xiaomi-mimo-v2.6-pro-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/xiaomi-mimo-v2.6-pro-free.0ce1bf03.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/xiaomi-mimo-v2.6-pro-free/llms.txt --- # GPT Realtime 2.1 Model id on AIHubMix: `gpt-realtime-2.1` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-Realtime-2.1 is a reasoning speech-to-speech model for the Realtime API, with tool use and configurable reasoning effort, improving on GPT-Realtime-2 in alphanumeric recognition, silence and noise handling, and interruption behavior. - Developer: OpenAI - Session kind: realtime WebSocket — speech-to-speech conversation - Input modalities: text, audio, image - Release date: 2026-07-06 - Pricing: see https://aihubmix.com/model/gpt-realtime-2.1 (realtime is billed on audio + text tokens) ## Endpoints (base URL: https://aihubmix.com) - `GET wss://aihubmix.com/v1/realtime?model=gpt-realtime-2.1` — realtime speech-to-speech conversation over WebSocket. Authenticate the handshake with `Authorization: Bearer $AIHUBMIX_API_KEY`. The model id is carried in the handshake query above (conversation sessions send no `intent`); audio is PCM16 / 24kHz / mono in both directions. ## Example ```python # pip install websockets import asyncio import base64 import json import os import websockets # Realtime conversation is a WebSocket session: model must be in the handshake URL. URL = "wss://aihubmix.com/v1/realtime?model=gpt-realtime-2.1" async def main(): # websockets >= 13 uses additional_headers; older versions use extra_headers async with websockets.connect( URL, additional_headers={"Authorization": "Bearer " + os.environ["AIHUBMIX_API_KEY"]} ) as ws: # 1) Configure the speech-to-speech session (voice + server VAD; no input transcription) await ws.send(json.dumps({ "type": "session.update", "session": { "type": "realtime", "audio": { "input": { "format": { "type": "audio/pcm", "rate": 24000 }, "turn_detection": { "type": "server_vad" } }, "output": { "format": { "type": "audio/pcm", "rate": 24000 }, "voice": "marin" } } } })) # 2) Stream your mic as raw PCM16 / 24kHz / mono in ~100ms chunks. Server VAD # detects when you stop talking and starts the reply automatically — # no commit / response.create needed. async def send_audio(): with open("audio_pcm16_24k.raw", "rb") as f: pcm = f.read() chunk = 24000 * 2 * 100 // 1000 # 100ms of 16-bit mono samples for i in range(0, len(pcm), chunk): await ws.send(json.dumps({ "type": "input_audio_buffer.append", "audio": base64.b64encode(pcm[i:i + chunk]).decode(), })) await asyncio.sleep(0.1) # simulate realtime pacing asyncio.create_task(send_audio()) # 3) Receive the reply: audio streams as base64 PCM16 (24kHz) — write it to a file you # can play; the transcript of what the model says arrives as text deltas. reply = open("assistant_reply_pcm16_24k.raw", "wb") async for msg in ws: evt = json.loads(msg) etype = evt.get("type", "") if etype == "input_audio_buffer.speech_started": # Barge-in: you started talking — stop/flush local playback here print("\n[listening…]") elif etype.endswith("audio_transcript.delta"): print(evt.get("delta", ""), end="", flush=True) elif etype.endswith("audio.delta"): reply.write(base64.b64decode(evt.get("delta", ""))) elif etype.endswith("response.done"): print("\n[reply complete]") elif etype == "error": print("\n[error]", evt.get("error")) break reply.close() asyncio.run(main()) ``` ## Response Realtime conversation is **speech-to-speech over one WebSocket** (audio in, audio out), not a request/response body. After the first `session.update` (which sets the voice and, deliberately, **no input transcription**), stream your mic as `input_audio_buffer.append` frames; server VAD starts the reply automatically — you do **not** send `response.create`. Read events until `response.done`: - `input_audio_buffer.speech_started` — the user began talking (flush local playback for barge-in) - `*.audio.delta` — assistant audio as base64 PCM16 / 24kHz (write it out and play it) - `*.audio_transcript.delta` — text transcript of what the assistant says - `response.done` — the assistant turn finished - `error` — a session-level error; the socket closes Note: the user's own speech is **not** transcribed on this channel (input transcription is off by design — enabling it closes the session). No HTTP response body. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-realtime-2.1 - Try in browser: https://aihubmix.com/playground?model=gpt-realtime-2.1 - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-realtime-2.1.9d8bc4a4.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (this realtime example was produced by its `generateRealtimeCode`; 7 languages, media and realtime included); the session frames it builds are the exact frames the gateway receives, so generated snippets and real requests cannot diverge - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-realtime-2.1/llms.txt --- # GLM 5.3 Model id on AIHubMix: `glm-5.3` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GLM-5.3 is Z.AI’s coding and agentic reasoning model, built for complex software engineering, long-running agent tasks, vulnerability analysis, and other demanding workloads. Building on GLM-5.2, it incorporates further post-training improvements to deliver stronger coding performance, better task execution, and greater token efficiency. We currently offer the production-ready GLM-5.3 API with unlimited concurrency, making it well suited for high-throughput workloads, coding agents, and large-scale automation. For a limited time, GLM-5.3 is available at 10% off. - Developer: Z.AI - Context window: 1,048,576 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2026-08-14 - Pricing: $1.1268/M input tokens, $3.9438/M output tokens, $0.2817/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"glm-5.3","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"glm-5.3","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/glm-5.3 - Try in browser: https://aihubmix.com/playground?model=glm-5.3 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/glm-5.3.a358b062.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/glm-5.3/llms.txt --- # GPT Image 2.5 Flare Model id on AIHubMix: `gpt-image-2.5-flare` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-Image-2.5 Flare is OpenAI's latest image model, the fastest and suited for everyday high-quality image generation. It accepts text and image inputs and produces image outputs. The model supports quality settings: low, medium, high, xhigh, max, and auto. - Developer: OpenAI - Input modalities: text, image - Release date: 2026-09-08 - Pricing: per-generation — see https://aihubmix.com/model/gpt-image-2.5-flare ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/edits` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-image-2.5-flare","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `image` | - | no | | | | `images` | array,null | no | | | | `mask` | string | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `output_format` | string,null | no | "png" | | | `prompt` | string | yes | | | | `size` | string|string|null | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/gpt-image-2.5-flare/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/gpt-image-2.5-flare/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/gpt-image-2.5-flare - Try in browser: https://aihubmix.com/playground?model=gpt-image-2.5-flare - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-image-2.5-flare/llms.txt --- # GPT Image 2.5 Sunburst Model id on AIHubMix: `gpt-image-2.5-sunburst` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT Image 2.5 Sunburst is OpenAI's latest image model, capable of generating and editing images from text and image inputs. It is suitable for workflows that require extremely high editing precision. The model supports quality settings: low, medium, high, xhigh, max, and auto. - Developer: OpenAI - Input modalities: text, image - Release date: 2026-09-08 - Pricing: per-generation — see https://aihubmix.com/model/gpt-image-2.5-sunburst ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/edits` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-image-2.5-sunburst","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `image` | - | no | | | | `images` | array,null | no | | | | `mask` | string | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `output_format` | string,null | no | "png" | | | `prompt` | string | yes | | | | `size` | string|string|null | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/gpt-image-2.5-sunburst/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/gpt-image-2.5-sunburst/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/gpt-image-2.5-sunburst - Try in browser: https://aihubmix.com/playground?model=gpt-image-2.5-sunburst - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-image-2.5-sunburst/llms.txt --- # Mercury 2.5 Preview Model id on AIHubMix: `mercury-2.5-preview` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Mercury 2.5 is the latest diffusion-based large language model (dLLM) released by Inception. It is the fastest inference LLM; unlike the sequential token-by-token generation approach, Mercury 2.5 can generate and optimize multiple tokens in parallel, achieving a generation speed of 1,107 tokens per second on standard GPUs. Compared to Mercury 2, its intelligence has increased by more than 10 percentage points, and its quality rivals leading cost-optimized frontier models such as GPT-5.6 Luna (Low), Gemini 3.5 Flash-Lite, and Claude Haiku 4.5. - Developer: Inception - Context window: 260,000 tokens - Input modalities: text - Capabilities: Tool calling, Structured outputs, Prompt caching - Release date: 2026-09-02 - Pricing: $0.2/M input tokens, $0.75/M output tokens, $0.02/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"mercury-2.5-preview","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"mercury-2.5-preview","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/mercury-2.5-preview - Try in browser: https://aihubmix.com/playground?model=mercury-2.5-preview - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/mercury-2.5-preview.4da604d5.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/mercury-2.5-preview/llms.txt --- # Qwen3.8 Max 2026 09-02 Model id on AIHubMix: `qwen3.8-max-2026-09-02` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Qwen3.8-Max-0902 (also known as qwen3.8-max-2026-09-02) is a snapshot version of Alibaba Cloud Tongyi Qianwen's qwen3.8-max. It pushes encoding depth further, enabling it to handle more complex engineering-level projects and long-term autonomous development; collaborative agent capabilities are significantly enhanced, performing more confidently in multi-tool orchestration and end-to-end delivery; visual understanding is comprehensively improved, with more sensitive and accurate chart reasoning, document parsing, and multimodal perception. Continuing the 1-million-context window, reasoning modes, and a complete tool ecosystem, it continues to evolve at a higher level of intelligence. - Developer: Qwen - Context window: 1,000,000 tokens - Input modalities: text, image, video - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Structured outputs, Prompt caching - Release date: 2026-09-02 - Pricing: $1.69/M input tokens, $5.07/M output tokens, $0.169/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen3.8-max-2026-09-02","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"qwen3.8-max-2026-09-02","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/qwen3.8-max-2026-09-02 - Try in browser: https://aihubmix.com/playground?model=qwen3.8-max-2026-09-02 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/qwen3.8-max-2026-09-02.e266d1de.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen3.8-max-2026-09-02/llms.txt --- # Coding GLM 5.3 Model id on AIHubMix: `coding-glm-5.3` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GLM-5.3 is Z.ai’s reasoning model for coding and agentic workflows, designed for complex software engineering, long-running agents, and vulnerability analysis. It uses the same base model as GLM-5.2, with scaled post-training improving coding, task execution, and token efficiency. This model is a limited-time preview version of GLM-5.3, intended for testing and evaluation only. Service stability is not guaranteed, and we do not recommend using it in production environments. We’re waiting for the official commercial API release and will integrate it as soon as official support becomes available. - Developer: Z.AI - Context window: 1,048,576 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2026-08-14 - Pricing: $0.06/M input tokens, $0.22/M output tokens, $0.015/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/messages` — Anthropic Messages (`x-api-key: $AIHUBMIX_API_KEY` + `anthropic-version: 2023-06-01`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"coding-glm-5.3","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"coding-glm-5.3","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/coding-glm-5.3 - Try in browser: https://aihubmix.com/playground?model=coding-glm-5.3 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/coding-glm-5.3.60aab1df.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/coding-glm-5.3/llms.txt --- # Coding GLM 5.3 Flash (free) Model id on AIHubMix: `coding-glm-5.3-flash-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > coding-glm-5.3-flash-free is the open and free version of coding-glm-5.3-flash. To ensure stable service performance, usage limits are in place: up to 5 requests per minute, 100 requests per day, and a daily token allowance of 1 million. - Developer: Z.AI - Context window: 1,000,000 tokens - Input modalities: text, image, video - Capabilities: Thinking, Streaming, Tool calling, Prompt caching - Release date: 2026-08-26 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"coding-glm-5.3-flash-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"coding-glm-5.3-flash-free","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/coding-glm-5.3-flash-free - Try in browser: https://aihubmix.com/playground?model=coding-glm-5.3-flash-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/coding-glm-5.3-flash-free.8ee3465e.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/coding-glm-5.3-flash-free/llms.txt --- # Hy4 Preview Model id on AIHubMix: `hy4-preview` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Hy4 Preview is Tencent Hunyuan’s large language model for agents, coding, office automation, and complex tool use.Hy4 preview has a total of 770B parameters and 49B active parameters, and is primarily optimized for agent, coding, and productivity scenarios. It strengthens understanding, planning, tool invocation, and sustained execution capabilities for complex tasks. Compared with the previous generation, Hy4 preview further improves multi-step agents, code development, and productivity tasks, offering better task decomposition, context continuity, instruction following, and long-horizon execution. In coding scenarios, it further enhances code understanding, generation, modification, and the handling of complex engineering tasks; in productivity scenarios, it focuses on improving document processing, information analysis, office automation, game development, webpage generation, and cross-tool collaboration. Hy4 preview is suitable for coding agents, complex tool invocation, and various agent workflows that require multi-step planning and sustained execution, providing more reliable task completion for complex real-world business scenarios. - Developer: Hunyuan - Context window: 1,048,576 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2026-08-28 - Pricing: $0.845/M input tokens, $2.535/M output tokens, $0.0423/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"hy4-preview","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"hy4-preview","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/hy4-preview - Try in browser: https://aihubmix.com/playground?model=hy4-preview - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/hy4-preview.70fffb4a.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/hy4-preview/llms.txt --- # Coding GLM 5.3 Flash Model id on AIHubMix: `coding-glm-5.3-flash` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Coding GLM 5.3 Flash is a dedicated version of GLM 5.3 Flash built for AI coding and Coding Agent workflows. It is designed for code understanding, generation, editing, repository-level development, and automated software engineering tasks. With support for context windows of up to approximately 1 million tokens, it can handle large codebases and extended development sessions. The model also supports text, image, and video inputs, along with tool use, making it well suited for AI coding tools such as Claude Code, OpenCode, Cline, and other agentic development environments. - Developer: Z.AI - Context window: 1,000,000 tokens - Input modalities: text, image, video - Capabilities: Thinking, Streaming, Tool calling, Prompt caching - Release date: 2026-08-26 - Pricing: $0.0282/M input tokens, $0.0986/M output tokens, $0.007/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"coding-glm-5.3-flash","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"coding-glm-5.3-flash","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/coding-glm-5.3-flash - Try in browser: https://aihubmix.com/playground?model=coding-glm-5.3-flash - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/coding-glm-5.3-flash.32a384b7.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/coding-glm-5.3-flash/llms.txt --- # Coding GLM 5.3 (free) Model id on AIHubMix: `coding-glm-5.3-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > coding-glm-5.3-free is the open and free version of coding-glm-5.3. To ensure stable service performance, usage limits are in place: up to 5 requests per minute, 100 requests per day, and a daily token allowance of 1 million. - Developer: Z.AI - Context window: 1,048,576 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2026-08-14 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"coding-glm-5.3-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"coding-glm-5.3-free","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/coding-glm-5.3-free - Try in browser: https://aihubmix.com/playground?model=coding-glm-5.3-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/coding-glm-5.3-free.e278e37e.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/coding-glm-5.3-free/llms.txt --- # Muse Spark 1.3 Model id on AIHubMix: `muse-spark-1.3` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Muse Spark 1.3 is Meta’s multimodal reasoning model, designed for long-running agentic, multi-agent, and complex coding workflows. It can maintain context and task constraints across extended workflows, reconcile conflicting information, and request clarification or confirmation when needed. Compared with Muse Spark 1.2, Muse Spark 1.3 improves efficiency across long-horizon agentic and coding tasks, with fewer unnecessary turns, tool calls, and tokens, while producing more concise outputs. - Developer: Meta - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio, PDF - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Citations, Prompt caching, Background mode, Server-side sessions - Release date: 2026-09-03 - Pricing: $1.375/M input tokens, $4.675/M output tokens, $0.165/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"muse-spark-1.3","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"muse-spark-1.3","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/muse-spark-1.3 - Try in browser: https://aihubmix.com/playground?model=muse-spark-1.3 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/muse-spark-1.3.20224cd2.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/muse-spark-1.3/llms.txt --- # Qwen3.8 Flash Model id on AIHubMix: `qwen3.8-flash` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Qwen3.8 Flash is Alibaba Cloud Qwen’s flagship native vision-language model for coding, office tasks, long-context reasoning, and agent workflows. It supports a 1M-token context, 128K output, web access, and tool calling. Compared with Qwen3.7-Plus, Qwen3.8-Flash significantly reduces training and inference costs—the training overhead is only about one-ninth of the former—while offering stronger capabilities on coding and office tasks. - Developer: Qwen - Context window: 1,000,000 tokens - Input modalities: text, image, video - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Structured outputs, Prompt caching - Release date: 2026-08-26 - Pricing: $0.1126/M input tokens, $0.38/M output tokens, $0.0141/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen3.8-flash","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"qwen3.8-flash","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/qwen3.8-flash - Try in browser: https://aihubmix.com/playground?model=qwen3.8-flash - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/qwen3.8-flash.ac8a0b24.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen3.8-flash/llms.txt --- # Gemini 3.7 Flash Model id on AIHubMix: `gemini-3.7-flash` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model is scheduled for retirement.** Calls still work until then. > > Once retired, requests return HTTP 404 `model_retired`. Responses already carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > (This model has been officially taken offline; requests to gemini-3.5-flash will now be automatically routed to gemini-3.8-flash.)Gemini 3.7 Flash is Google’s natively multimodal reasoning model for coding, agents, web development, and knowledge work. It supports a 1M-token context window and adjustable thinking levels. Compared with Gemini 3.6 Flash, it improves coding, tool use, multi-step planning, and instruction following. - Developer: Google - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio, PDF - Capabilities: Thinking, Streaming, Tool calling, Web search, URL context, Code interpreter, Computer use, File search, Structured outputs, Prompt caching - Release date: 2026-08-13 - Pricing: $0.75/M input tokens, $3.75/M output tokens, $0.075/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /gemini/v1beta/models/gemini-3.7-flash:generateContent` — Google Gemini (`x-goog-api-key: $AIHUBMIX_API_KEY`; streaming: `:streamGenerateContent?alt=sse`) - `POST /v1/messages` — Anthropic Messages (`x-api-key: $AIHUBMIX_API_KEY` + `anthropic-version: 2023-06-01`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-3.7-flash","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gemini-3.7-flash","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gemini-3.7-flash - Try in browser: https://aihubmix.com/playground?model=gemini-3.7-flash - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gemini-3.7-flash.ddc06a8b.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-3.7-flash/llms.txt --- # Mai Image 2.6 Model id on AIHubMix: `mai-image-2.6` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > MAI-Image-2.6 is Microsoft's latest image-generation and editing model. It can generate images from text prompts and supports image-guided editing in multiple aspect ratios. - Developer: Microsoft - Input modalities: text, image - Release date: 2026-09-05 - Pricing: per-generation — see https://aihubmix.com/model/mai-image-2.6 ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/edits` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"mai-image-2.6","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `image` | - | no | | | | `model` | string | yes | | | | `output_format` | string,null | no | "png" | png | | `prompt` | string | yes | | | | `resolution` | string | no | | 512, 1K, 2K, 4K | | `response_format` | string,null | no | "url" | url, b64_json, null | | `size` | string,null | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/mai-image-2.6/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/mai-image-2.6/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/mai-image-2.6 - Try in browser: https://aihubmix.com/playground?model=mai-image-2.6 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/mai-image-2.6/llms.txt --- # Mai Image 2.6 Flash Model id on AIHubMix: `mai-image-2.6-flash` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > MAI-Image-2.6 Flash is Microsoft’s low-latency version of the latest image model MAI-Image-2.6. It supports image generation in multiple aspect ratios and image-guided editing. - Developer: Microsoft - Input modalities: text, image - Release date: 2026-09-05 - Pricing: per-generation — see https://aihubmix.com/model/mai-image-2.6-flash ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/edits` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"mai-image-2.6-flash","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `image` | - | no | | | | `model` | string | yes | | | | `output_format` | string,null | no | "png" | png | | `prompt` | string | yes | | | | `resolution` | string | no | | 512, 1K, 2K, 4K | | `response_format` | string,null | no | "url" | url, b64_json, null | | `size` | string,null | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/mai-image-2.6-flash/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/mai-image-2.6-flash/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/mai-image-2.6-flash - Try in browser: https://aihubmix.com/playground?model=mai-image-2.6-flash - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/mai-image-2.6-flash/llms.txt --- # Xiaomi Mimo V2.6 Flash (free) Model id on AIHubMix: `xiaomi-mimo-v2.6-flash-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > xiaomi-mimo-v2.6-flash-free is the open free version of xiaomi-mimo-v2.6-flash. To maintain reliable service, each account is limited to 5 requests per minute, 100 requests per day, and 1 million tokens per day. - Developer: Xiaomi - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Release date: 2026-09-22 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"xiaomi-mimo-v2.6-flash-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"xiaomi-mimo-v2.6-flash-free","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/xiaomi-mimo-v2.6-flash-free - Try in browser: https://aihubmix.com/playground?model=xiaomi-mimo-v2.6-flash-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/xiaomi-mimo-v2.6-flash-free.d95bff9a.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/xiaomi-mimo-v2.6-flash-free/llms.txt --- # Agnes 3.0 Flash Model id on AIHubMix: `agnes-3.0-flash` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Agnes 3.0 Flash is designed for real-world agent tasks and development workflows, covering the full execution chain from task understanding and planning to tool invocation and final delivery. The model focuses on improving stability, instruction-following, factual grounding, and output completeness in complex tasks, helping developers build more reliable agent applications. - Developer: Agnes - Context window: 512,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling - Release date: 2026-09-09 - Pricing: $0.03/M input tokens, $0.15/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/responses` — OpenAI Responses (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"agnes-3.0-flash","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"agnes-3.0-flash","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/agnes-3.0-flash - Try in browser: https://aihubmix.com/playground?model=agnes-3.0-flash - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/agnes-3.0-flash.c5c8d1f2.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/agnes-3.0-flash/llms.txt --- # MiniMax H3 Max Model id on AIHubMix: `minimax-h3-max` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > MiniMax H3 (minimax-h3) is a general-purpose omni-modal generation model developed by the Chinese AI company MiniMax. It can jointly understand text, images, video, and audio while generating videos of up to 15 seconds in 2K resolution with native stereo sound. It is designed for advertising, branding, e-commerce, product design, UI/UX, and gaming. Compared with Hailuo 01 and Hailuo 02, H3 evolves from specialized video generation into a unified model for multimodal creation, reference-based editing, text rendering, and motion transfer. - Developer: Minimax - Input modalities: text, image - Release date: 2026-09-02 - Pricing: per-generation — see https://aihubmix.com/model/minimax-h3-max ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /ai/v1/videos/{id}` until a terminal state ## Example ```bash curl -s https://aihubmix.com/ai/v1/videos \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"minimax-h3-max","prompt":""}' ``` The response contains a task id — poll `GET https://aihubmix.com/ai/v1/videos/` until `status` is terminal, then read the output URL(s). Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (video endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string|string|null | no | "16:9" | | | `duration` | integer,null | no | 5 | | | `extra` | object,null | no | | | | `frame_images` | array,null | no | | | | `input_references` | array,null | no | | | | `model` | string | yes | | | | `prompt` | string | yes | | | | `resolution` | string|null | no | "768P" | | | `webhook_url` | string,null | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/minimax-h3-max/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/minimax-h3-max/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/minimax-h3-max - Try in browser: https://aihubmix.com/playground?model=minimax-h3-max - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/minimax-h3-max/llms.txt --- # GLM 5.2 Model id on AIHubMix: `glm-5.2` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GLM-5.2 is Z.ai’s flagship model for the era of long-horizon tasks. With a truly usable 1M-token context window, it can handle project-level engineering context, execute long-running tasks more reliably, follow engineering standards more consistently, and complete the full development workflow from requirements to multi-platform deployment in a single task. - Developer: Z.AI - Context window: 1,000,000 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2026-06-16 - Pricing: $1.1268/M input tokens, $3.9438/M output tokens, $0.2817/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/messages` — Anthropic Messages (`x-api-key: $AIHUBMIX_API_KEY` + `anthropic-version: 2023-06-01`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"glm-5.2","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"glm-5.2","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/glm-5.2 - Try in browser: https://aihubmix.com/playground?model=glm-5.2 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/glm-5.2.8b06b4f8.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/glm-5.2/llms.txt --- # DeepSeek V4 Flash 0731 Model id on AIHubMix: `deepseek-v4-flash-0731` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > DeepSeek-V4-Flash-0731(deepseek-v4-flash-0731) is an open-source MoE large language model developed by the Chinese AI company DeepSeek, with support for a million-token context window. It is designed for coding, complex reasoning, tool use, agentic workflows, and long-document processing. Its advantages include strong performance with fewer active parameters and improved efficiency through DSpark speculative decoding. Compared with DeepSeek V4-Flash Preview, it offers significantly stronger coding and agent capabilities, while outperforming DeepSeek V4-Pro Preview on several benchmarks with fewer active parameters. - Developer: DeepSeek - Context window: 1,000,000 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2026-07-31 - Pricing: $0.142/M input tokens, $0.284/M output tokens, $0.0284/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/messages` — Anthropic Messages (`x-api-key: $AIHUBMIX_API_KEY` + `anthropic-version: 2023-06-01`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"deepseek-v4-flash-0731","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"deepseek-v4-flash-0731","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/deepseek-v4-flash-0731 - Try in browser: https://aihubmix.com/playground?model=deepseek-v4-flash-0731 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/deepseek-v4-flash-0731.b3a859ff.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/deepseek-v4-flash-0731/llms.txt --- # DeepSeek V4 Flash Vision Exp Model id on AIHubMix: `deepseek-v4-flash-vision-exp` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > DeepSeek’s officially released new multimodal visual-understanding model, DeepSeek‑V4‑Flash‑Vision‑Exp, is experimental in nature and supports multimodal inputs. In pure-text capabilities (agents, reasoning, world knowledge, etc.), DeepSeek‑V4‑Flash‑Vision‑Exp is on par with the official DeepSeek‑V4‑Flash release.(The model name deepseek-v4-flash-vision-exp remains callable, but the official corresponding model has been taken offline; requests will be served by the DeepSeek-V4.1-Flash model and billed at Flash pricing.) - Developer: DeepSeek - Context window: 1,000,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2026-07-31 - Pricing: $0.155/M input tokens, $0.62/M output tokens, $0.0031/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/messages` — Anthropic Messages (`x-api-key: $AIHUBMIX_API_KEY` + `anthropic-version: 2023-06-01`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"deepseek-v4-flash-vision-exp","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"deepseek-v4-flash-vision-exp","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/deepseek-v4-flash-vision-exp - Try in browser: https://aihubmix.com/playground?model=deepseek-v4-flash-vision-exp - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/deepseek-v4-flash-vision-exp.55a94915.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/deepseek-v4-flash-vision-exp/llms.txt --- # DeepSeek V4 Pro 0813 Model id on AIHubMix: `deepseek-v4-pro-0813` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > DeepSeek V4 Pro 0813 is DeepSeek’s high-performance general-purpose reasoning and agent model, designed for complex reasoning, coding, long-document analysis, and agentic workflows. It supports thinking and non-thinking modes, a 1M-token context window, up to 384K output, tool calling, and the Responses API. Compared with V4 Flash 0731, Pro prioritizes capability on complex tasks, while Flash focuses on speed, cost efficiency, and high concurrency. - Developer: DeepSeek - Context window: 1,000,000 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Release date: 2026-08-13 - Pricing: $0.6918/M input tokens, $2.0754/M output tokens, $0.023/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"deepseek-v4-pro-0813","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"deepseek-v4-pro-0813","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/deepseek-v4-pro-0813 - Try in browser: https://aihubmix.com/playground?model=deepseek-v4-pro-0813 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/deepseek-v4-pro-0813.bf96d058.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/deepseek-v4-pro-0813/llms.txt --- # Dots 3 Note Preview (free) Model id on AIHubMix: `dots-3-note-preview-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Dots3-Note Preview is an open-weight mixture-of-experts model developed by Dots Studio, featuring 16B active parameters out of 280B total. As the lightest model in the Dots 3 family, it is designed for efficient performance while supporting an expansive context length of 512,000 tokens. This preview version provides an accessible way to experience the capabilities of the Dots 3 architecture. - Developer: Dots Studio - Context window: 524,288 tokens - Input modalities: text, image, video, audio - Capabilities: Thinking, Tool calling, Document input - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"dots-3-note-preview-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"dots-3-note-preview-free","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/dots-3-note-preview-free - Try in browser: https://aihubmix.com/playground?model=dots-3-note-preview-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/dots-3-note-preview-free.f3306a46.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/dots-3-note-preview-free/llms.txt --- # Grok 4.6 Model id on AIHubMix: `grok-4.6` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Grok 4.6 is xAI’s (SpaceXAI) flagship multimodal reasoning model for coding, long-running agents, knowledge work, and interactive application development. It supports image understanding, a 500K context window, tool calling, and structured outputs. Compared with Grok 4.5, it offers stronger multi-step execution, self-verification, coding, and visual project generation. - Developer: Grok - Context window: 500,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Document input, Structured outputs, Prompt caching, Server-side sessions - Release date: 2026-08 - Pricing: $2/M input tokens, $6/M output tokens, $0.5/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"grok-4.6","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"grok-4.6","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/grok-4.6 - Try in browser: https://aihubmix.com/playground?model=grok-4.6 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/grok-4.6.06150b6b.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/grok-4.6/llms.txt --- # Coding Xiaomi Mimo V2.6 Flash Model id on AIHubMix: `coding-xiaomi-mimo-v2.6-flash` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Only supports OpenAI-compatible formats. - Developer: Xiaomi - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Release date: 2026-09-22 - Pricing: $0.08/M input tokens, $0.16/M output tokens, $0.0016/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"coding-xiaomi-mimo-v2.6-flash","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"coding-xiaomi-mimo-v2.6-flash","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/coding-xiaomi-mimo-v2.6-flash - Try in browser: https://aihubmix.com/playground?model=coding-xiaomi-mimo-v2.6-flash - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/coding-xiaomi-mimo-v2.6-flash.945fb83c.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/coding-xiaomi-mimo-v2.6-flash/llms.txt --- # DeepSeek V4 Flash 0731 Fast Model id on AIHubMix: `deepseek-v4-flash-0731-fast` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > DeepSeek V4 Flash 0731 Fast is a high-speed deployment of DeepSeek’s agentic model provided by Wafer, designed for coding, tool use, and high-volume agent workloads. It preserves the capabilities of V4 Flash 0731 while delivering much faster inference. Compared with V4 Pro, it prioritizes latency and execution efficiency. - Developer: DeepSeek - Context window: 1,000,000 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2026-07-31 - Pricing: $0.28/M input tokens, $1.4/M output tokens, $0.07/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/messages` — Anthropic Messages (`x-api-key: $AIHUBMIX_API_KEY` + `anthropic-version: 2023-06-01`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"deepseek-v4-flash-0731-fast","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"deepseek-v4-flash-0731-fast","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/deepseek-v4-flash-0731-fast - Try in browser: https://aihubmix.com/playground?model=deepseek-v4-flash-0731-fast - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/deepseek-v4-flash-0731-fast.eae86f83.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/deepseek-v4-flash-0731-fast/llms.txt --- # Mai Thinking 1 Model id on AIHubMix: `mai-thinking-1` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > MAI-Thinking-1 is Microsoft’s first inference model in the MAI series, built for enterprise-scale workloads. With excellent reasoning, mathematical, and general intelligence capabilities, combined with superior cost-effectiveness, it makes high-throughput, 24/7 AI workloads economically viable. - Developer: Microsoft - Context window: 256,000 tokens - Input modalities: text - Capabilities: Streaming, Tool calling, Prompt caching - Release date: 2026-08-12 - Pricing: $2/M input tokens, $8/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"mai-thinking-1","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"mai-thinking-1","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/mai-thinking-1 - Try in browser: https://aihubmix.com/playground?model=mai-thinking-1 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/mai-thinking-1.be34ae72.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/mai-thinking-1/llms.txt --- # Wan3.0 Video Model id on AIHubMix: `wan3.0-video` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Wan 3.0 (Tongyi Wanxiang 3.0) is an integrated video generation and editing model launched by Alibaba’s Tongyi Lab. It is suitable for scenarios such as advertising, e-commerce, short films, character animation, video editing, and document-to-video. Its advantages include native support for videos up to 30 seconds long and the unified handling of reference content such as text, images, audio, video, documents, spreadsheets, presentations, and web pages, delivering more realistic visuals and audio, stable character consistency, and capabilities for referencing, editing, replication, and driving. Compared with Wan 2.7, which split text-to-video, image-to-video, reference generation, and editing across multiple models, Wan 3.0 integrates these capabilities into a single model, making the creative workflow more unified and flexible. - Developer: Qwen - Input modalities: text, image, video, audio, document, link - Pricing: per-generation — see https://aihubmix.com/model/wan3.0-video ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /ai/v1/videos/{id}` until a terminal state - `POST /v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /v1/videos/{id}` until a terminal state ## Example ```bash curl -s https://aihubmix.com/ai/v1/videos \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"wan3.0-video"}' ``` The response contains a task id — poll `GET https://aihubmix.com/ai/v1/videos/` until `status` is terminal, then read the output URL(s). Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (video endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string|string|null | no | "adaptive" | | | `duration` | integer|null | no | 5 | -1, 2, 3, 4, 5, 6, … | | `extra` | object,null | no | | | | `frame_images` | array | no | | | | `generate_audio` | boolean,null | no | true | | | `input_references` | array | no | | | | `model` | string | yes | | | | `prompt` | string | no | | | | `resolution` | string|null | no | "1080p" | | | `seed` | integer,null | no | | | | `webhook_url` | string,null | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/wan3.0-video/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/wan3.0-video/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/wan3.0-video - Try in browser: https://aihubmix.com/playground?model=wan3.0-video - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/wan3.0-video/llms.txt --- # Wan3.0 Video Prime Model id on AIHubMix: `wan3.0-video-prime` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Wan3.0 Video Prime is Alibaba Cloud’s preview high-speed edition of its All-in-One video model, positioned for faster-turnaround ads, ecommerce, short films, and editing. Aligned with Wan3.0 Video’s capabilities, it supports text-to-video, image-to-video, reference-based generation, document/webpage-to-video, up to 30 seconds at 30fps, 1080P output, adaptive aspect ratios, and native audio. It unifies modes that Wan 2.7 handled through separate task-specific models. - Developer: Qwen - Input modalities: text, image, video, audio, document, link - Pricing: per-generation — see https://aihubmix.com/model/wan3.0-video-prime ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /ai/v1/videos/{id}` until a terminal state - `POST /v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /v1/videos/{id}` until a terminal state ## Example ```bash curl -s https://aihubmix.com/ai/v1/videos \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"wan3.0-video-prime"}' ``` The response contains a task id — poll `GET https://aihubmix.com/ai/v1/videos/` until `status` is terminal, then read the output URL(s). Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (video endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string|string|null | no | "adaptive" | | | `duration` | integer|null | no | 5 | -1, 2, 3, 4, 5, 6, … | | `extra` | object,null | no | | | | `frame_images` | array | no | | | | `generate_audio` | boolean,null | no | true | | | `input_references` | array | no | | | | `model` | string | yes | | | | `prompt` | string | no | | | | `resolution` | string|null | no | "1080p" | | | `seed` | integer,null | no | | | | `webhook_url` | string,null | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/wan3.0-video-prime/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/wan3.0-video-prime/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/wan3.0-video-prime - Try in browser: https://aihubmix.com/playground?model=wan3.0-video-prime - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/wan3.0-video-prime/llms.txt --- # GPT 5.6 Sol Disc Model id on AIHubMix: `gpt-5.6-sol-disc` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-5.6 Sol (limited-time 50% off) is OpenAI’s frontier reasoning model for complex coding, professional knowledge work, deep research, and long-running agents. It supports a roughly 1.05M-token context window, image understanding, and extensive tool use. Compared with Terra and Luna, Sol prioritizes capability and reliability on demanding tasks. - Developer: OpenAI - Context window: 1,050,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching, Server-side sessions - Release date: 2026-08-18 - Pricing: $4/M input tokens, $20/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/responses` — OpenAI Responses (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-5.6-sol-disc","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-5.6-sol-disc","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-5.6-sol-disc - Try in browser: https://aihubmix.com/playground?model=gpt-5.6-sol-disc - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-5.6-sol-disc.153da0ea.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-5.6-sol-disc/llms.txt --- # Doubao Seedance 2.5 260628 Model id on AIHubMix: `doubao-seedance-2-5-260628` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Seedance 2.5 is ByteDance Seed’s next-generation unified multimodal audio-video generation model, designed for filmmaking, advertising, education, simulation, and long-form content creation. It generates up to 30 seconds of synchronized audio and video in one pass and supports multi-round extensions. Compared with Seedance 2.0, it offers stronger storytelling, smoother transitions, richer reference support, improved realism, and more precise editing. - Developer: ByteDance - Input modalities: text, image, video, audio - Release date: 2026-06-28 - Pricing: per-generation — see https://aihubmix.com/model/doubao-seedance-2-5-260628 ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /ai/v1/videos/{id}` until a terminal state ## Example ```bash curl -s https://aihubmix.com/ai/v1/videos \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"doubao-seedance-2-5-260628","prompt":""}' ``` The response contains a task id — poll `GET https://aihubmix.com/ai/v1/videos/` until `status` is terminal, then read the output URL(s). Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (video endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string|string|null | no | "adaptive" | | | `duration` | - | no | -1 | | | `extra` | object,null | no | | | | `frame_images` | array,null | no | | | | `generate_audio` | boolean,null | no | true | | | `input_references` | array,null | no | | | | `model` | string | yes | | | | `prompt` | string | yes | | | | `resolution` | string|null | no | "720p" | | | `webhook_url` | string,null | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/doubao-seedance-2-5-260628/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/doubao-seedance-2-5-260628/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/doubao-seedance-2-5-260628 - Try in browser: https://aihubmix.com/playground?model=doubao-seedance-2-5-260628 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/doubao-seedance-2-5-260628/llms.txt --- # GPT 5.6 Luna Model id on AIHubMix: `gpt-5.6-luna` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-5.6 Luna is designed for cost-sensitive, high-volume workloads. It roughly corresponds to the nano model tier used in earlier GPT-5 families. - Developer: OpenAI - Context window: 1,050,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching, Server-side sessions - Release date: 2026-07-09 - Pricing: $0.2/M input tokens, $1.2/M output tokens, $0.02/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/responses` — OpenAI Responses (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-5.6-luna","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-5.6-luna","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-5.6-luna - Try in browser: https://aihubmix.com/playground?model=gpt-5.6-luna - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-5.6-luna.aaef42f0.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-5.6-luna/llms.txt --- # GPT 5.6 Sol Model id on AIHubMix: `gpt-5.6-sol` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT‑5.6 Sol sets a new standard for both intelligence and efficiency, achieving state-of-the-art results across coding, knowledge work, cybersecurity, and science while outperforming previous and competing frontier models with fewer tokens and at lower estimated cost. - Developer: OpenAI - Context window: 1,050,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching, Server-side sessions - Release date: 2026-07-09 - Pricing: $4/M input tokens, $20/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/responses` — OpenAI Responses (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-5.6-sol","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-5.6-sol","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-5.6-sol - Try in browser: https://aihubmix.com/playground?model=gpt-5.6-sol - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-5.6-sol.98f0d914.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-5.6-sol/llms.txt --- # GPT 5.6 Terra Model id on AIHubMix: `gpt-5.6-terra` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-5.6 Terra is designed for workloads that balance intelligence and cost. It roughly corresponds to the mini model tier used in earlier GPT-5 families. - Developer: OpenAI - Context window: 1,050,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching, Server-side sessions - Release date: 2026-07-09 - Pricing: $2/M input tokens, $12/M output tokens, $0.2/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/responses` — OpenAI Responses (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-5.6-terra","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-5.6-terra","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-5.6-terra - Try in browser: https://aihubmix.com/playground?model=gpt-5.6-terra - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-5.6-terra.5a4aabfd.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-5.6-terra/llms.txt --- # Qwen3.8 Max Model id on AIHubMix: `qwen3.8-max` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Qwen3.8-Max is Alibaba Cloud Tongyi Qianwen's next-generation flagship large language model, featuring a mixture-of-experts (MoE) architecture with 2.4 trillion parameters. It achieves another breakthrough in encoding depth, enabling it to handle more complex engineering-grade projects and long-term autonomous development; collaborative agent capabilities are significantly enhanced, performing more confidently in multi-tool orchestration and end-to-end delivery; visual understanding is comprehensively improved, with more sensitive and accurate chart reasoning, document parsing, and multimodal perception. Continuing the 1-million-context window, reasoning modes, and a complete tool ecosystem, it continues to evolve at a higher level of intelligence. - Developer: Qwen - Context window: 1,000,000 tokens - Input modalities: text, image, video - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Structured outputs, Prompt caching - Release date: 2026-08-03 - Pricing: $1.69/M input tokens, $5.07/M output tokens, $0.169/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen3.8-max","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"qwen3.8-max","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/qwen3.8-max - Try in browser: https://aihubmix.com/playground?model=qwen3.8-max - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/qwen3.8-max.7695c787.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen3.8-max/llms.txt --- # Agnes 2.5 Flash Model id on AIHubMix: `agnes-2.5-flash` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Agnes 2.5 Flash is Agnes AI’s fast and efficient language model, an upgraded, fully available model based on Agnes 2.0 Flash. It continues to use an OpenAI-compatible Chat Completions interface and has been optimized for coding tasks, agent workflows, tool calls, multi-turn dialogue, reasoning, and image understanding. - Developer: Agnes - Context window: 512,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Prompt caching - Release date: 2026-07 - Pricing: $0.03/M input tokens, $0.15/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/responses` — OpenAI Responses (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"agnes-2.5-flash","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"agnes-2.5-flash","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/agnes-2.5-flash - Try in browser: https://aihubmix.com/playground?model=agnes-2.5-flash - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/agnes-2.5-flash.269abf4f.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/agnes-2.5-flash/llms.txt --- # Agnes 2.5 Pro Model id on AIHubMix: `agnes-2.5-pro` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Agnes 2.5 Pro is Agnes AI’s paid inference model and the commercially stable version of the Agnes 2.5 Pro Alpha ranking model, suitable for advanced coding, scientific reasoning, long-context analysis, agent workflows, and multimodal understanding. The model is accessed via an OpenAI-compatible Chat Completions API. - Developer: Agnes - Context window: 1,000,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Prompt caching - Release date: 2026-08-01 - Pricing: $0.45/M input tokens, $0.9/M output tokens, $0.0038/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/responses` — OpenAI Responses (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"agnes-2.5-pro","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"agnes-2.5-pro","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/agnes-2.5-pro - Try in browser: https://aihubmix.com/playground?model=agnes-2.5-pro - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/agnes-2.5-pro.5ddcca4b.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/agnes-2.5-pro/llms.txt --- # Agnes 2.5 Pro Alpha Model id on AIHubMix: `agnes-2.5-pro-alpha` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Agnes 2.5 Pro Alpha is Agnes AI’s paid inference model, suitable for advanced coding, scientific reasoning, long-context analysis, agent workflows, and multimodal understanding. The model is accessed via an OpenAI-compatible Chat Completions API. - Developer: Agnes - Context window: 1,000,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Prompt caching - Release date: 2026-07-24 - Pricing: $0.45/M input tokens, $0.9/M output tokens, $0.0038/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/responses` — OpenAI Responses (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"agnes-2.5-pro-alpha","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"agnes-2.5-pro-alpha","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/agnes-2.5-pro-alpha - Try in browser: https://aihubmix.com/playground?model=agnes-2.5-pro-alpha - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/agnes-2.5-pro-alpha.c08863a6.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/agnes-2.5-pro-alpha/llms.txt --- # Agnes Image 2.1 Flash Model id on AIHubMix: `agnes-image-2.1-flash` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Agnes Image 2.1 Flash is Agnes AI’s high-performance image generation and image editing model, supporting text-to-image, image-to-image, and multi-image composition. It is suitable for creative design, marketing visuals, e-commerce product images, and social content production. - Developer: Agnes - Input modalities: text, image - Pricing: per-generation — see https://aihubmix.com/model/agnes-image-2.1-flash ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/edits` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"agnes-image-2.1-flash","prompt":"","size":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string|null | no | "1:1" | | | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `image` | - | no | | | | `images` | array,null | no | | | | `model` | string | yes | | | | `prompt` | string | yes | | | | `response_format` | string,null | no | "url" | url, b64_json, null | | `size` | string|string|null | yes | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/agnes-image-2.1-flash/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/agnes-image-2.1-flash/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/agnes-image-2.1-flash - Try in browser: https://aihubmix.com/playground?model=agnes-image-2.1-flash - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/agnes-image-2.1-flash/llms.txt --- # Grok 4.5 Model id on AIHubMix: `grok-4.5` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Grok 4.5 was trained on datasets spanning knowledge in coding, science, engineering, and math. With both intelligent and efficient reasoning, Grok 4.5 excels at real engineering tasks and exceeds comparable leading models at these tasks. - Developer: Grok - Context window: 500,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Document input, Structured outputs, Prompt caching, Server-side sessions - Release date: 2026-07 - Pricing: $2/M input tokens, $6/M output tokens, $0.5/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"grok-4.5","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"grok-4.5","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/grok-4.5 - Try in browser: https://aihubmix.com/playground?model=grok-4.5 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/grok-4.5.56b0b4ed.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/grok-4.5/llms.txt --- # Lfm 2.5 2.6b (free) Model id on AIHubMix: `lfm-2.5-2.6b-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > LFM-2.5-2.6B is a compact reasoning model developed by Liquid, featuring a generous 128,000 token context length. It is highly suited for agent workflows, data extraction, retrieval-augmented generation (RAG), and long-context processing. However, the developer advises against using this model for agentic coding tasks. - Developer: Liquid - Context window: 131,072 tokens - Input modalities: text - Capabilities: Thinking, Tool calling, Structured outputs - Release date: 2026-08-12 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"lfm-2.5-2.6b-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"lfm-2.5-2.6b-free","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/lfm-2.5-2.6b-free - Try in browser: https://aihubmix.com/playground?model=lfm-2.5-2.6b-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/lfm-2.5-2.6b-free.274e4d5c.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/lfm-2.5-2.6b-free/llms.txt --- # MiniMax H3 Model id on AIHubMix: `minimax-h3` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > MiniMax H3 (minimax-h3) is a general-purpose omni-modal generation model developed by the Chinese AI company MiniMax. It can jointly understand text, images, video, and audio while generating videos of up to 15 seconds in 2K resolution with native stereo sound. It is designed for advertising, branding, e-commerce, product design, UI/UX, and gaming. Compared with Hailuo 01 and Hailuo 02, H3 evolves from specialized video generation into a unified model for multimodal creation, reference-based editing, text rendering, and motion transfer. - Developer: Minimax - Input modalities: text, image, video, audio - Release date: 2026-07-30 - Pricing: per-generation — see https://aihubmix.com/model/minimax-h3 ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /ai/v1/videos/{id}` until a terminal state ## Example ```bash curl -s https://aihubmix.com/ai/v1/videos \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"minimax-h3","prompt":""}' ``` The response contains a task id — poll `GET https://aihubmix.com/ai/v1/videos/` until `status` is terminal, then read the output URL(s). Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (video endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string|string|null | no | "16:9" | | | `duration` | integer|null | no | 5 | | | `extra` | object,null | no | | | | `frame_images` | array,null | no | | | | `input_references` | array,null | no | | | | `model` | string | yes | | | | `prompt` | string | yes | | | | `resolution` | string|null | no | "768P" | | | `webhook_url` | string,null | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/minimax-h3/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/minimax-h3/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/minimax-h3 - Try in browser: https://aihubmix.com/playground?model=minimax-h3 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/minimax-h3/llms.txt --- # Qwen3.8 2.4t A95B Model id on AIHubMix: `qwen3.8-2.4t-a95b` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Qwen3.8-2.4T-A95B is Alibaba’s most powerful Qwen model to date. It is a 2.4‑trillion‑parameter sparse Mixture-of-Experts (MoE) model with approximately 95 billion active parameters. It is built for autonomous, long‑duration tasks: multi‑day code runs, reproducing research papers, and self‑improvement. - Developer: Qwen - Context window: 1,000,000 tokens - Input modalities: text - Capabilities: Thinking, Tool calling, Web search, Code interpreter, Structured outputs, Prompt caching - Release date: 2026-08-12 - Pricing: $2/M input tokens, $6/M output tokens, $0.5/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen3.8-2.4t-a95b","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"qwen3.8-2.4t-a95b","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/qwen3.8-2.4t-a95b - Try in browser: https://aihubmix.com/playground?model=qwen3.8-2.4t-a95b - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/qwen3.8-2.4t-a95b.9cad84d3.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen3.8-2.4t-a95b/llms.txt --- # Claude Opus 5 Model id on AIHubMix: `claude-opus-5` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Claude Opus 5 is Anthropic’s flagship model for demanding reasoning, coding, and long-horizon agentic work. It is particularly strong at end-to-end software tasks, code review and bug finding, visual analysis of charts and documents, complex office deliverables, and coordinating parallel subagents. The model maintains strong instruction following and tool use across extended tasks, while remaining effective at lower effort settings for workloads that prioritize latency and token efficiency. - Developer: Anthropic - Context window: 1,000,000 tokens - Input modalities: text, image, PDF - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Computer use, Memory tool, Structured outputs, Citations, Prompt caching - Release date: 2026-07-24 - Pricing: $5/M input tokens, $25/M output tokens, $0.5/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/messages` — Anthropic Messages (`x-api-key: $AIHUBMIX_API_KEY` + `anthropic-version: 2023-06-01`) - `POST /v1/responses` — OpenAI Responses (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"claude-opus-5","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"claude-opus-5","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/claude-opus-5 - Try in browser: https://aihubmix.com/playground?model=claude-opus-5 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/claude-opus-5.8231dfa2.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/claude-opus-5/llms.txt --- # Gemini 3.6 Flash Model id on AIHubMix: `gemini-3.6-flash` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Gemini 3.6 Flash provides sustained frontier-level intelligence optimized for real-world tasks at a higher speed and lower cost. Designed for the agentic era, it excels at code generation, agentic execution, and spatial reasoning. This model is particularly effective for rapid agentic loops involving complex coding cycles and iterations. - Developer: Google - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio, PDF - Capabilities: Thinking, Tool calling, Web search, URL context, Code interpreter, Computer use, File search, Structured outputs, Prompt caching - Release date: 2026-07-21 - Pricing: $0.75/M input tokens, $3.75/M output tokens, $0.075/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /gemini/v1beta/models/gemini-3.6-flash:generateContent` — Google Gemini (`x-goog-api-key: $AIHUBMIX_API_KEY`; streaming: `:streamGenerateContent?alt=sse`) - `POST /v1/messages` — Anthropic Messages (`x-api-key: $AIHUBMIX_API_KEY` + `anthropic-version: 2023-06-01`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-3.6-flash","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gemini-3.6-flash","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gemini-3.6-flash - Try in browser: https://aihubmix.com/playground?model=gemini-3.6-flash - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gemini-3.6-flash.2cd7b148.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-3.6-flash/llms.txt --- # Nemotron 3.5 Lightning (free) Model id on AIHubMix: `nemotron-3.5-lightning-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > NVIDIA Nemotron 3.5 Lightning is an open mixture-of-experts model from NVIDIA, featuring 3 billion active parameters out of 30 billion total. It supports an extensive context window of up to 1,000,000 tokens. This model is designed to deliver efficient performance for high-throughput agentic workloads and specialized tasks. Compared with Nemotron 3 Super and Ultra, Lightning is smaller and optimized for fast, high-volume execution, while the larger models focus on complex planning and advanced reasoning. - Developer: Nvidia - Context window: 1,048,576 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Structured outputs - Release date: 2026-08-11 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"nemotron-3.5-lightning-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"nemotron-3.5-lightning-free","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/nemotron-3.5-lightning-free - Try in browser: https://aihubmix.com/playground?model=nemotron-3.5-lightning-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/nemotron-3.5-lightning-free.879fb4e8.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/nemotron-3.5-lightning-free/llms.txt --- # Qwen Image 3.0 Model id on AIHubMix: `qwen-image-3.0` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Qwen Image 3.0(qwen-image-3.0) is an image generation and editing model developed by Alibaba Cloud’s Qwen team. It supports text-to-image generation, reference-based creation, and image editing. It is well suited for social media, e-commerce, creative design, and everyday content production, offering a strong balance of image quality and speed. Compared with the Pro version, it is better suited for frequent and large-scale daily creation. - Developer: Qwen - Input modalities: text, image - Pricing: per-generation — see https://aihubmix.com/model/qwen-image-3.0 ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen-image-3.0","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `image` | - | no | | | | `images` | array,null | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `negative_prompt` | string,null | no | | | | `output_format` | string,null | no | "png" | png | | `prompt` | string | yes | | | | `response_format` | string,null | no | "url" | url, null | | `seed` | integer,null | no | | | | `size` | string|null | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/qwen-image-3.0/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/qwen-image-3.0/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/qwen-image-3.0 - Try in browser: https://aihubmix.com/playground?model=qwen-image-3.0 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen-image-3.0/llms.txt --- # Qwen Image 3.0 Pro Model id on AIHubMix: `qwen-image-3.0-pro` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Qwen Image 3.0 Pro (qwen-image-3.0-pro) is Alibaba Cloud Qwen’s flagship image generation and editing model. It is designed for advertising, brand visuals, UI, presentations, product imagery, and professional design. Its strengths include complex layouts, accurate Chinese and English text rendering, realistic materials, and reference-based editing. Compared with the standard version, it delivers stronger detail, composition, and commercial-grade visual quality. - Developer: Qwen - Input modalities: text, image - Pricing: per-generation — see https://aihubmix.com/model/qwen-image-3.0-pro ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen-image-3.0-pro","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `image` | - | no | | | | `images` | array,null | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `negative_prompt` | string,null | no | | | | `output_format` | string,null | no | "png" | png | | `prompt` | string | yes | | | | `response_format` | string,null | no | "url" | url, null | | `seed` | integer,null | no | | | | `size` | string|null | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/qwen-image-3.0-pro/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/qwen-image-3.0-pro/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/qwen-image-3.0-pro - Try in browser: https://aihubmix.com/playground?model=qwen-image-3.0-pro - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen-image-3.0-pro/llms.txt --- # Claude Sonnet 5 Model id on AIHubMix: `claude-sonnet-5` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Claude Sonnet 5 is the next generation of Anthropic's Sonnet model family. It is a drop-in upgrade for Claude Sonnet 4.6 with three behavior changes: adaptive thinking is on by default, manual extended thinking now returns a 400 error (it was deprecated on Claude Sonnet 4.6), and setting sampling parameters (temperature, top_p, top_k) to non-default values returns a 400 error - Developer: Anthropic - Context window: 1,000,000 tokens - Input modalities: text, image, PDF - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Computer use, Memory tool, Structured outputs, Citations, Prompt caching - Release date: 2026-06-30 - Pricing: $2/M input tokens, $10/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /gemini/v1beta/models/claude-sonnet-5:generateContent` — Google Gemini (`x-goog-api-key: $AIHUBMIX_API_KEY`; streaming: `:streamGenerateContent?alt=sse`) - `POST /v1/messages` — Anthropic Messages (`x-api-key: $AIHUBMIX_API_KEY` + `anthropic-version: 2023-06-01`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"claude-sonnet-5","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"claude-sonnet-5","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/claude-sonnet-5 - Try in browser: https://aihubmix.com/playground?model=claude-sonnet-5 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/claude-sonnet-5.3305d225.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/claude-sonnet-5/llms.txt --- # Kimi K3 Model id on AIHubMix: `kimi-k3` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Kimi K3 is Kimi’s flagship model for long-horizon coding and end-to-end knowledge work, with a 1M-token context window and industry-leading intelligence. - Developer: Moonshot AI - Context window: 1,048,576 tokens - Input modalities: text, image, video - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2026-07-16 - Pricing: $2.84/M input tokens, $14.2/M output tokens, $0.284/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"kimi-k3","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"kimi-k3","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/kimi-k3 - Try in browser: https://aihubmix.com/playground?model=kimi-k3 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/kimi-k3.ab6aabc2.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/kimi-k3/llms.txt --- # Muse Spark 1.2 Model id on AIHubMix: `muse-spark-1.2` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Muse Spark 1.1 is a multimodal reasoning model from Meta, built for agentic tasks. It accepts text, images, video, audio, and PDF documents and returns text, with a 1M-token context window. - Developer: Meta - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio, PDF - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Citations, Prompt caching, Background mode, Server-side sessions - Release date: 2026-08-05 - Pricing: $1.375/M input tokens, $4.675/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"muse-spark-1.2","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"muse-spark-1.2","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/muse-spark-1.2 - Try in browser: https://aihubmix.com/playground?model=muse-spark-1.2 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/muse-spark-1.2.5bb78dc4.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/muse-spark-1.2/llms.txt --- # Gemini 3.1 Flash Lite Image Model id on AIHubMix: `gemini-3.1-flash-lite-image` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Google's newest, most compact, and most cost-effective image generation and editing model, designed for large-scale use. - Developer: Google - Input modalities: text, image - Release date: 2026-06-30 - Pricing: per-generation — see https://aihubmix.com/model/gemini-3.1-flash-lite-image ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/edits` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-3.1-flash-lite-image","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string|null | no | null | | | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `image` | - | no | | | | `images` | array,null | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `prompt` | string | yes | | | | `resolution` | string | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/gemini-3.1-flash-lite-image/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/gemini-3.1-flash-lite-image/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/gemini-3.1-flash-lite-image - Try in browser: https://aihubmix.com/playground?model=gemini-3.1-flash-lite-image - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-3.1-flash-lite-image/llms.txt --- # Gemini 3.5 Flash Lite Model id on AIHubMix: `gemini-3.5-flash-lite` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Gemini 3.5 Flash-Lite is a low-latency, cost-effective multimodal model optimized for high-throughput, low-cost execution for subagent tasks and document parsing. The model supports text, image, video, audio, and PDF inputs, and is designed for high-volume agentic workflows, simple data extraction, and applications where latency and API cost are the primary constraints. - Developer: Google - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio, PDF - Capabilities: Thinking, Tool calling, Web search, URL context, Code interpreter, File search, Structured outputs, Prompt caching - Release date: 2026-07-21 - Pricing: $0.3/M input tokens, $2.5/M output tokens, $0.03/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /gemini/v1beta/models/gemini-3.5-flash-lite:generateContent` — Google Gemini (`x-goog-api-key: $AIHUBMIX_API_KEY`; streaming: `:streamGenerateContent?alt=sse`) - `POST /v1/messages` — Anthropic Messages (`x-api-key: $AIHUBMIX_API_KEY` + `anthropic-version: 2023-06-01`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-3.5-flash-lite","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gemini-3.5-flash-lite","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gemini-3.5-flash-lite - Try in browser: https://aihubmix.com/playground?model=gemini-3.5-flash-lite - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gemini-3.5-flash-lite.6e5ea4ca.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-3.5-flash-lite/llms.txt --- # GLM 5.2 Fast Preview Model id on AIHubMix: `glm-5.2-fast-preview` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GLM-5.2-Fast-Preview is the high-speed version of Zhipu AI’s flagship model GLM-5.2, supporting a 1M ultra-long context. The model’s capabilities are aligned with the GLM-5.2 standard version, offering logical reasoning, long-text understanding, and code generation. Through inference acceleration optimizations, output TPS can reach 1.5–2× that of the GLM-5.2 standard version, significantly improving output speed. It is suitable for scenarios sensitive to output speed, such as real-time dialogue, multi-turn Agent calls, and streaming code generation. - Developer: Z.AI - Context window: 1,000,000 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2026-06-16 - Pricing: $2.254/M input tokens, $7.889/M output tokens, $0.5635/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/messages` — Anthropic Messages (`x-api-key: $AIHUBMIX_API_KEY` + `anthropic-version: 2023-06-01`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"glm-5.2-fast-preview","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"glm-5.2-fast-preview","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/glm-5.2-fast-preview - Try in browser: https://aihubmix.com/playground?model=glm-5.2-fast-preview - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/glm-5.2-fast-preview.e0ef17f6.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/glm-5.2-fast-preview/llms.txt --- # Muse Spark 1.1 Model id on AIHubMix: `muse-spark-1.1` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Muse Spark 1.1 is a multimodal reasoning model from Meta, built for agentic tasks. It accepts text, images, video, audio, and PDF documents and returns text, with a 1M-token context window. - Developer: Meta - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio, PDF - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Citations, Prompt caching, Background mode, Server-side sessions - Release date: 2026-04-08 - Pricing: $1.375/M input tokens, $4.675/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"muse-spark-1.1","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"muse-spark-1.1","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/muse-spark-1.1 - Try in browser: https://aihubmix.com/playground?model=muse-spark-1.1 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/muse-spark-1.1.c1452217.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/muse-spark-1.1/llms.txt --- # Claude Fable 5 Model id on AIHubMix: `claude-fable-5` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Anthropic's most capable widely released model, for the most demanding reasoning and long-horizon agentic work(This model is extremely expensive and is not recommended for casual use.) - Developer: Anthropic - Context window: 1,000,000 tokens - Input modalities: text, image, PDF - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Memory tool, Structured outputs, Citations, Prompt caching - Release date: 2026-06-09 - Pricing: $11/M input tokens, $55/M output tokens, $1.1/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/messages` — Anthropic Messages (`x-api-key: $AIHUBMIX_API_KEY` + `anthropic-version: 2023-06-01`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"claude-fable-5","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"claude-fable-5","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/claude-fable-5 - Try in browser: https://aihubmix.com/playground?model=claude-fable-5 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/claude-fable-5.83f81302.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/claude-fable-5/llms.txt --- # Claude Opus 4.8 Model id on AIHubMix: `claude-opus-4-8` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Claude Opus 4.8 is Anthropic’s newest and most powerful publicly available model. It is suitable for the most complex tasks. It is Anthropic’s strongest model for complex reasoning, long-horizon agent programming, and highly autonomous work.claude-opus-4-8 does not display thought content by default; you need to set the additional parameter "display": "summarized" to enable it - Developer: Anthropic - Context window: 1,000,000 tokens - Input modalities: text, image, PDF - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Computer use, Memory tool, Structured outputs, Citations, Prompt caching - Release date: 2026-05-28 - Pricing: $5/M input tokens, $25/M output tokens, $0.5/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/messages` — Anthropic Messages (`x-api-key: $AIHUBMIX_API_KEY` + `anthropic-version: 2023-06-01`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"claude-opus-4-8","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"claude-opus-4-8","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/claude-opus-4-8 - Try in browser: https://aihubmix.com/playground?model=claude-opus-4-8 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/claude-opus-4-8.97fd8748.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/claude-opus-4-8/llms.txt --- # Hy3 Model id on AIHubMix: `hy3` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > The Hy3 official version is honed for real-world business scenarios, using a Mixture-of-Experts (MoE) architecture with 295B total parameters and 21B activated parameters. It natively supports a 256K context window and offers multiple thinking modes: no_think (ultra-fast response), think_low (quick thinking), and think_high (deep reasoning), balancing ultra-fast responses, complex reasoning, and invocation cost. Compared with the Preview version, Hy3—based on real business feedback from Tencent Yuanbao, WorkBuddy, ima, Marvis, and others—focuses on improving the Coding Agent, long-form understanding, multi-turn context continuity, search QA, and complex task execution, performing more stably in reducing hallucinations, improving task completion, and engineering usability. It is better suited to practical scenarios such as frontend tasks, cross-file code development, long-document analysis, office automation, and multi-step Agent workflows. - Developer: Hunyuan - Context window: 256,000 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2026-07-06 - Pricing: $0.1562/M input tokens, $0.6248/M output tokens, $0.0391/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"hy3","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"hy3","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/hy3 - Try in browser: https://aihubmix.com/playground?model=hy3 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/hy3.2fc8e923.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/hy3/llms.txt --- # Doubao Seed 2.1 Pro Model id on AIHubMix: `doubao-seed-2-1-pro` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > A new generation of large models moving toward production-grade intelligence, comprehensively upgrading coding, agent, and multimodal capabilities with stronger autonomous planning, long-horizon execution, and dynamic recovery abilities to handle real, complex enterprise tasks. - Developer: ByteDance - Context window: 256,000 tokens - Input modalities: text, image, video, document - Capabilities: Thinking, Streaming, Tool calling, Web search, Document input, Structured outputs, Prompt caching, Server-side sessions - Pricing: $0.9295/M input tokens, $4.6475/M output tokens, $0.1859/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"doubao-seed-2-1-pro","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"doubao-seed-2-1-pro","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/doubao-seed-2-1-pro - Try in browser: https://aihubmix.com/playground?model=doubao-seed-2-1-pro - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/doubao-seed-2-1-pro.fd5068e6.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/doubao-seed-2-1-pro/llms.txt --- # Doubao Seed 2.1 Turbo Model id on AIHubMix: `doubao-seed-2-1-turbo` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Balancing performance and cost, comprehensively upgrading coding, agent, and multimodal capabilities, with stronger autonomous planning, long-horizon execution, and dynamic recovery abilities to handle real, complex enterprise tasks. - Developer: ByteDance - Context window: 256,000 tokens - Input modalities: text, image, video, document - Capabilities: Thinking, Tool calling, Web search, Document input, Structured outputs, Prompt caching, Server-side sessions - Pricing: $0.4647/M input tokens, $2.3237/M output tokens, $0.093/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"doubao-seed-2-1-turbo","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"doubao-seed-2-1-turbo","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/doubao-seed-2-1-turbo - Try in browser: https://aihubmix.com/playground?model=doubao-seed-2-1-turbo - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/doubao-seed-2-1-turbo.e30c14e9.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/doubao-seed-2-1-turbo/llms.txt --- # Mai Image 2.5 Pro Model id on AIHubMix: `mai-image-2.5-pro` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > MAI-Image-2.5 is Microsoft's flagship AI image generation and editing model. With top-tier realism, precise text rendering, and powerful image editing capabilities, it debuted among the top three in AI image generation rankings. It is primarily targeted at commercial design, product photography, and professional creative work. - Developer: Microsoft - Input modalities: text, image - Pricing: per-generation — see https://aihubmix.com/model/mai-image-2.5-pro ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"mai-image-2.5-pro","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `images` | array,null | no | | | | `model` | string | yes | | | | `output_format` | string,null | no | "png" | png | | `prompt` | string | yes | | | | `response_format` | string,null | no | "url" | url, b64_json, null | | `size` | string,null | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/mai-image-2.5-pro/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/mai-image-2.5-pro/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/mai-image-2.5-pro - Try in browser: https://aihubmix.com/playground?model=mai-image-2.5-pro - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/mai-image-2.5-pro/llms.txt --- # Gemini 3.5 Flash Model id on AIHubMix: `gemini-3.5-flash` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model is scheduled for retirement.** Calls still work until then. > > Once retired, requests return HTTP 404 `model_retired`. Responses already carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > (This model has been officially taken offline; requests to gemini-3.5-flash will now be automatically routed to gemini-3.6-flash.)Gemini 3.5 Flash provides sustained frontier-level intelligence optimized for real-world tasks at a higher speed and lower cost. Designed for the agentic era, it excels at sub-agent deployment, multi-step workflows, and long-horizon tasks at scale. This model is particularly effective for rapid agentic loops involving complex coding cycles and iterations. - Developer: Google - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio, PDF - Capabilities: Thinking, Tool calling, Web search, URL context, Code interpreter, Computer use, File search, Structured outputs, Prompt caching - Release date: 2026-05-19 - Pricing: $0.75/M input tokens, $3.75/M output tokens, $0.075/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /gemini/v1beta/models/gemini-3.5-flash:generateContent` — Google Gemini (`x-goog-api-key: $AIHUBMIX_API_KEY`; streaming: `:streamGenerateContent?alt=sse`) - `POST /v1/messages` — Anthropic Messages (`x-api-key: $AIHUBMIX_API_KEY` + `anthropic-version: 2023-06-01`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-3.5-flash","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gemini-3.5-flash","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gemini-3.5-flash - Try in browser: https://aihubmix.com/playground?model=gemini-3.5-flash - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gemini-3.5-flash.a884b9c7.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-3.5-flash/llms.txt --- # Grok Build 0.1 Model id on AIHubMix: `grok-build-0.1` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Fast coding model trained specifically for agentic coding workflows. - Developer: Grok - Context window: 256,000 tokens - Input modalities: text, image - Capabilities: Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2026-04-16 - Pricing: $1/M input tokens, $2/M output tokens, $0.2/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"grok-build-0.1","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"grok-build-0.1","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/grok-build-0.1 - Try in browser: https://aihubmix.com/playground?model=grok-build-0.1 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/grok-build-0.1.83d5729d.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/grok-build-0.1/llms.txt --- # Mai Image 2.5 Model id on AIHubMix: `mai-image-2.5` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > MAI-Image-2.5 is Microsoft's flagship AI image generation and editing model. With top-tier realism, precise text rendering, and powerful image editing capabilities, it debuted among the top three in AI image generation rankings. It is primarily targeted at commercial design, product photography, and professional creative work. - Developer: Microsoft - Input modalities: text, image - Pricing: per-generation — see https://aihubmix.com/model/mai-image-2.5 ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/edits` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"mai-image-2.5","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `image` | string | no | | | | `model` | string | yes | | | | `output_format` | string,null | no | "png" | png | | `prompt` | string | yes | | | | `response_format` | string,null | no | "url" | url, b64_json, null | | `size` | string,null | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/mai-image-2.5/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/mai-image-2.5/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/mai-image-2.5 - Try in browser: https://aihubmix.com/playground?model=mai-image-2.5 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/mai-image-2.5/llms.txt --- # Mai Image 2.5 Flash Model id on AIHubMix: `mai-image-2.5-flash` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > MAI-Image-2.5 is Microsoft's flagship AI image generation and editing model. With top-tier realism, precise text rendering, and powerful image editing capabilities, it debuted among the top three in AI image generation rankings. It is primarily targeted at commercial design, product photography, and professional creative work. - Developer: Microsoft - Input modalities: text, image - Pricing: per-generation — see https://aihubmix.com/model/mai-image-2.5-flash ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"mai-image-2.5-flash","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `images` | array,null | no | | | | `model` | string | yes | | | | `output_format` | string,null | no | "png" | png | | `prompt` | string | yes | | | | `response_format` | string,null | no | "url" | url, b64_json, null | | `size` | string,null | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/mai-image-2.5-flash/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/mai-image-2.5-flash/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/mai-image-2.5-flash - Try in browser: https://aihubmix.com/playground?model=mai-image-2.5-flash - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/mai-image-2.5-flash/llms.txt --- # Coding Kimi K3 Model id on AIHubMix: `coding-kimi-k3` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms - Developer: Moonshot AI - Context window: 1,048,576 tokens - Input modalities: text, image, video - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2026-07-16 - Pricing: $0.44/M input tokens, $1.6133/M output tokens, $0.066/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"coding-kimi-k3","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"coding-kimi-k3","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/coding-kimi-k3 - Try in browser: https://aihubmix.com/playground?model=coding-kimi-k3 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/coding-kimi-k3.002014fe.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/coding-kimi-k3/llms.txt --- # Happyhorse 1.1 I2v Model id on AIHubMix: `happyhorse-1.1-i2v` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > HappyHorse-1.1-I2V supports image-to-video generation, further enhancing visual texture, dynamic performance, and cross-segment consistency. The model can more accurately understand input images and carry forward the creative intent, delivering significant improvements in character skin texture, ID consistency across segments, motion smoothness, text rendering stability, and audio–visual synchronization, producing higher-quality videos that are more realistic and natural, rich in detail, and more consistent. - Developer: Qwen - Input modalities: text, image - Release date: 2026-07-17 - Pricing: per-generation — see https://aihubmix.com/model/happyhorse-1.1-i2v ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /ai/v1/videos/{id}` until a terminal state - `POST /v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /v1/videos/{id}` until a terminal state ## Example ```bash curl -s https://aihubmix.com/ai/v1/videos \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"happyhorse-1.1-i2v","prompt":""}' ``` The response contains a task id — poll `GET https://aihubmix.com/ai/v1/videos/` until `status` is terminal, then read the output URL(s). Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (video endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `duration` | integer|null | no | 5 | | | `extra` | object,null | no | | | | `frame_images` | array,null | no | | | | `generate_audio` | boolean,null | no | | | | `model` | string | yes | | | | `prompt` | string | yes | | | | `resolution` | string,null | no | "720p" | 720p, 1080p, null | | `seed` | integer,null | no | | | | `size` | string,null | no | | | | `webhook_url` | string,null | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/happyhorse-1.1-i2v/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/happyhorse-1.1-i2v/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/happyhorse-1.1-i2v - Try in browser: https://aihubmix.com/playground?model=happyhorse-1.1-i2v - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/happyhorse-1.1-i2v/llms.txt --- # Happyhorse 1.1 R2v Model id on AIHubMix: `happyhorse-1.1-r2v` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > HappyHorse-1.1-R2V supports reference-based video generation, further improving the stable preservation of subject, scene style, and visual consistency. The model supports up to nine reference images as input, and can more accurately understand and carry forward creative intent, offering stronger controllability and expressiveness in characters, scenes, style, and camera/shot performance. - Developer: Qwen - Input modalities: text, image - Release date: 2026-07-17 - Pricing: per-generation — see https://aihubmix.com/model/happyhorse-1.1-r2v ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /ai/v1/videos/{id}` until a terminal state - `POST /v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /v1/videos/{id}` until a terminal state ## Example ```bash curl -s https://aihubmix.com/ai/v1/videos \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"happyhorse-1.1-r2v","prompt":""}' ``` The response contains a task id — poll `GET https://aihubmix.com/ai/v1/videos/` until `status` is terminal, then read the output URL(s). Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (video endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string,null | no | "16:9" | 16:9, 9:16, 1:1, 4:3, 3:4, null | | `duration` | integer|null | no | 5 | | | `extra` | object,null | no | | | | `frame_images` | array,null | no | | | | `generate_audio` | boolean,null | no | | | | `input_references` | array,null | no | | | | `model` | string | yes | | | | `prompt` | string | yes | | | | `resolution` | string,null | no | "720p" | 720p, 1080p, null | | `seed` | integer,null | no | | | | `size` | string,null | no | | | | `webhook_url` | string,null | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/happyhorse-1.1-r2v/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/happyhorse-1.1-r2v/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/happyhorse-1.1-r2v - Try in browser: https://aihubmix.com/playground?model=happyhorse-1.1-r2v - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/happyhorse-1.1-r2v/llms.txt --- # Happyhorse 1.1 T2v Model id on AIHubMix: `happyhorse-1.1-t2v` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > HappyHorse-1.1-T2V supports text-to-video generation, further enhancing text semantic understanding, camera/shot orchestration, and dynamic generation performance. The model can more accurately realize creative intent, producing higher-quality videos with smoother, more natural movements, richer detail, and greater consistency in character actions, scene atmosphere, visual aesthetics, and physical motion. - Developer: Qwen - Input modalities: text - Release date: 2026-07-17 - Pricing: per-generation — see https://aihubmix.com/model/happyhorse-1.1-t2v ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /ai/v1/videos/{id}` until a terminal state - `POST /v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /v1/videos/{id}` until a terminal state ## Example ```bash curl -s https://aihubmix.com/ai/v1/videos \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"happyhorse-1.1-t2v","prompt":""}' ``` The response contains a task id — poll `GET https://aihubmix.com/ai/v1/videos/` until `status` is terminal, then read the output URL(s). Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (video endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string,null | no | "16:9" | 16:9, 9:16, 1:1, 4:3, 3:4, null | | `duration` | integer|null | no | 5 | | | `extra` | object,null | no | | | | `generate_audio` | boolean,null | no | | | | `model` | string | yes | | | | `prompt` | string | yes | | | | `resolution` | string,null | no | "720p" | 720p, 1080p, null | | `seed` | integer,null | no | | | | `size` | string,null | no | | | | `webhook_url` | string,null | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/happyhorse-1.1-t2v/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/happyhorse-1.1-t2v/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/happyhorse-1.1-t2v - Try in browser: https://aihubmix.com/playground?model=happyhorse-1.1-t2v - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/happyhorse-1.1-t2v/llms.txt --- # Coding GLM 5.2 (free) Model id on AIHubMix: `coding-glm-5.2-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > coding glm 5.2 free (coding-glm-5.2-free) is a free coding-focused API route offered by AIHubMix, powered by Z.ai’s open-source flagship GLM-5.2 model. It is suitable for code generation, debugging, project-level development, tool use, and long-running agent tasks, with support for reasoning, function calling, and structured outputs. Compared with GLM-5.1, it offers stronger coding capabilities and a 1-million-token context window. Each account is limited to 5 requests per minute, 100 requests per day, and 1 million free tokens per day. - Developer: Z.AI - Context window: 1,000,000 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2026-06-16 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"coding-glm-5.2-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"coding-glm-5.2-free","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/coding-glm-5.2-free - Try in browser: https://aihubmix.com/playground?model=coding-glm-5.2-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/coding-glm-5.2-free.af2ce9b2.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/coding-glm-5.2-free/llms.txt --- # Coding Kimi K3 (free) Model id on AIHubMix: `coding-kimi-k3-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > coding-kimi-k3-free is the open and free version of coding-kimi-k3. To maintain reliable service, each account is limited to 5 requests per minute, 100 requests per day, and 1 million tokens per day. - Developer: Moonshot AI - Context window: 1,048,576 tokens - Input modalities: text, image, video - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2026-07-16 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"coding-kimi-k3-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"coding-kimi-k3-free","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/coding-kimi-k3-free - Try in browser: https://aihubmix.com/playground?model=coding-kimi-k3-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/coding-kimi-k3-free.f169c095.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/coding-kimi-k3-free/llms.txt --- # Gemini 3.1 Flash Image Model id on AIHubMix: `gemini-3.1-flash-image` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > gemini-3.1-flash-image (Nano Banana 2) features professional-grade visual intelligence, lightning-fast efficiency, and realistic, grounded generative capabilities. This model serves as the high-efficiency counterpart to Gemini 3 Pro Image, optimized for speed and high-volume developer use cases. - Developer: Google - Input modalities: text, image, PDF, video - Release date: 2026-05-28 - Pricing: per-generation — see https://aihubmix.com/model/gemini-3.1-flash-image ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/edits` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-3.1-flash-image","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string|null | no | null | | | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `image` | - | no | | | | `images` | array,null | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `prompt` | string | yes | | | | `resolution` | string | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/gemini-3.1-flash-image/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/gemini-3.1-flash-image/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/gemini-3.1-flash-image - Try in browser: https://aihubmix.com/playground?model=gemini-3.1-flash-image - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-3.1-flash-image/llms.txt --- # Kimi K2.7 Code Model id on AIHubMix: `kimi-k2.7-code` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Kimi K2.7 Code is Kimi’s most intelligent Coding model, capable of completing programming tasks with higher success rates in long context. It features a native multimodal architecture that supports text, image, video input, and thinking modes, and dialogue and agent tasks. - Developer: Moonshot AI - Context window: 262,144 tokens - Input modalities: text, image, video - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2026-06-12 - Pricing: $0.95/M input tokens, $3.9995/M output tokens, $0.1608/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"kimi-k2.7-code","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"kimi-k2.7-code","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/kimi-k2.7-code - Try in browser: https://aihubmix.com/playground?model=kimi-k2.7-code - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/kimi-k2.7-code.5a09c66c.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/kimi-k2.7-code/llms.txt --- # Kimi K2.7 Code Highspeed Model id on AIHubMix: `kimi-k2.7-code-highspeed` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > High-Speed version of Kimi K2.7 Code model, with output speed of approximately 180 Tokens/s and up to 260 Tokens/s in short context scenarios, delivering a more extreme coding experience. - Developer: Moonshot AI - Context window: 262,144 tokens - Input modalities: text, image, video - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2026-06-12 - Pricing: $1.9/M input tokens, $7.999/M output tokens, $0.3217/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"kimi-k2.7-code-highspeed","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"kimi-k2.7-code-highspeed","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/kimi-k2.7-code-highspeed - Try in browser: https://aihubmix.com/playground?model=kimi-k2.7-code-highspeed - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/kimi-k2.7-code-highspeed.ab6342e8.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/kimi-k2.7-code-highspeed/llms.txt --- # Gemini 3 Pro Image Model id on AIHubMix: `gemini-3-pro-image` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Gemini-3-Pro-Image (Nano Banana Pro) is a high-performance image generation and editing model built on Gemini 3 Pro. It delivers enhanced multimodal understanding and real-world semantic reasoning, enabling fast creation of well-structured visual content such as infographics, product sketches, and multi-subject scenes. It can also leverage real-time knowledge through Search grounding. The model excels in text rendering, consistent multi-image blending, and identity preservation, while offering fine-grained creative controls like localized edits, lighting and focus adjustments, camera transformations, and flexible aspect ratios. It’s ideal for rapid design, concept previews, product visualization, and everyday image generation workflows. - Developer: Google - Input modalities: text, image - Release date: 2026-05-28 - Pricing: per-generation — see https://aihubmix.com/model/gemini-3-pro-image ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/edits` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-3-pro-image","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string|null | no | null | | | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `image` | - | no | | | | `images` | array,null | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `prompt` | string | yes | | | | `resolution` | string | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/gemini-3-pro-image/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/gemini-3-pro-image/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/gemini-3-pro-image - Try in browser: https://aihubmix.com/playground?model=gemini-3-pro-image - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-3-pro-image/llms.txt --- # GPT Audio 1.5 Model id on AIHubMix: `gpt-audio-1.5` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > The gpt-audio model is OpenAI's first officially released (generally available) audio model. It supports audio input and output and can be used in Chat Completions. - Developer: OpenAI - Context window: 128,000 tokens - Input modalities: text, audio - Capabilities: Streaming, Tool calling, Audio output, Multimodal output - Pricing: $2.5/M input tokens, $10/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-audio-1.5","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-audio-1.5","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-audio-1.5 - Try in browser: https://aihubmix.com/playground?model=gpt-audio-1.5 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-audio-1.5.e05ba781.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-audio-1.5/llms.txt --- # Kling V3 Omni Model id on AIHubMix: `kling-v3-omni` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > VIDEO 3.0 Omni: All-in-One Multimodal Input, Voice-Driven Characters, Direct Audio-Visual Output, and Storyboarding Building on the Kling VIDEO O1 and Kling VIDEO 2.6, the Kling 3.0 Model Series leverage a deeply integrated unified model training framework, achieving more native multimodal input and output. It combines Native Audio with Element Consistency Control, and breaks through duration limits. - Developer: KLing - Input modalities: text, image, video - Pricing: per-generation — see https://aihubmix.com/model/kling-v3-omni ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /ai/v1/videos/{id}` until a terminal state - `POST /v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /v1/videos/{id}` until a terminal state ## Example ```bash curl -s https://aihubmix.com/ai/v1/videos \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"kling-v3-omni","prompt":""}' ``` The response contains a task id — poll `GET https://aihubmix.com/ai/v1/videos/` until `status` is terminal, then read the output URL(s). Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (video endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string,null | no | "16:9" | 16:9, 9:16, 1:1, 4:3, 3:4, 3:2, … | | `duration` | integer|null | no | 5 | | | `extra` | object,null | no | | | | `frame_images` | array,null | no | | | | `generate_audio` | boolean,null | no | | | | `input_references` | array,null | no | | | | `model` | string | yes | | | | `prompt` | string | yes | | | | `resolution` | string,null | no | | 480p, 720p, 1080p, 1K, 2K, 4K, … | | `seed` | integer,null | no | | | | `size` | string,null | no | | | | `webhook_url` | string,null | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/kling-v3-omni/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/kling-v3-omni/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/kling-v3-omni - Try in browser: https://aihubmix.com/playground?model=kling-v3-omni - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/kling-v3-omni/llms.txt --- # Kling Video O1 Model id on AIHubMix: `kling-video-o1` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Kling Video O1 is a major unified multimodal video model launched by Kuaishou. It integrates text, images, video, and subject inputs into a single engine, focusing on addressing the "consistency" pain point in AI video, and is very suitable for professional teams in e-commerce, advertising, and continuity-driven creation that need to maintain the appearance of characters and objects unchanged across scenes. - Developer: KLing - Input modalities: text, image, video - Pricing: per-generation — see https://aihubmix.com/model/kling-video-o1 ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /ai/v1/videos/{id}` until a terminal state - `POST /v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /v1/videos/{id}` until a terminal state ## Example ```bash curl -s https://aihubmix.com/ai/v1/videos \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"kling-video-o1","prompt":""}' ``` The response contains a task id — poll `GET https://aihubmix.com/ai/v1/videos/` until `status` is terminal, then read the output URL(s). Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (video endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string,null | no | "16:9" | 16:9, 9:16, 1:1, 4:3, 3:4, 3:2, … | | `duration` | integer|null | no | 5 | | | `extra` | object,null | no | | | | `frame_images` | array,null | no | | | | `generate_audio` | boolean,null | no | | | | `input_references` | array,null | no | | | | `model` | string | yes | | | | `prompt` | string | yes | | | | `resolution` | string,null | no | | 480p, 720p, 1080p, 1K, 2K, 4K, … | | `seed` | integer,null | no | | | | `size` | string,null | no | | | | `webhook_url` | string,null | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/kling-video-o1/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/kling-video-o1/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/kling-video-o1 - Try in browser: https://aihubmix.com/playground?model=kling-video-o1 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/kling-video-o1/llms.txt --- # Longcat 2.0 Model id on AIHubMix: `longcat-2.0` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Designed for agent development scenarios, it natively supports tool invocation, multi-step reasoning, and long-context tasks; it excels at code generation, automated workflows, and executing complex instructions; and it is deeply adapted to productivity tools such as Claude Code, OpenClaw, OpenCode, and Kilo Code. - Developer: Meituan - Context window: 1,000,000 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Prompt caching - Release date: 2026-06-30 - Pricing: $0.7746/M input tokens, $3.0984/M output tokens, $0.0155/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"longcat-2.0","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"longcat-2.0","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/longcat-2.0 - Try in browser: https://aihubmix.com/playground?model=longcat-2.0 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/longcat-2.0.e7625ad0.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/longcat-2.0/llms.txt --- # MiniMax M3 Model id on AIHubMix: `minimax-m3` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > The MiniMax M3 is a flagship programming model built for real-world productivity. As a production-grade model natively designed for Agent scenarios, it has achieved state-of-the-art (SOTA) performance in coding, agentic tool use, search, and office work. - Developer: Minimax - Context window: 1,000,000 tokens - Input modalities: text, image, video - Capabilities: Thinking, Streaming, Tool calling, Web search, Prompt caching - Release date: 2026-06-01 - Pricing: $0.288/M input tokens, $1.152/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"minimax-m3","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"minimax-m3","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/minimax-m3 - Try in browser: https://aihubmix.com/playground?model=minimax-m3 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/minimax-m3.c04378f7.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/minimax-m3/llms.txt --- # Qwen3.7 Flash Model id on AIHubMix: `qwen3.7-flash` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > The Qwen 3.7 series' mid-to-high cost-performance "Plus" model builds on strong text capabilities with a comprehensive upgrade to vision-language abilities, while retaining full agent capabilities in coding, tool use, and productivity workflows. Its core features are multimodal interactive hybrid agent capabilities, able to perceive real-world scenes, read screens and operate GUIs, generate code based on visual references, and provide end-to-end navigation of mobile applications. - Developer: Qwen - Context window: 1,000,000 tokens - Input modalities: text, image, video - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Structured outputs, Prompt caching - Release date: 2026-07-15 - Pricing: $0.0282/M input tokens, $0.1128/M output tokens, $0.0056/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen3.7-flash","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"qwen3.7-flash","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/qwen3.7-flash - Try in browser: https://aihubmix.com/playground?model=qwen3.7-flash - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/qwen3.7-flash.a8b81288.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen3.7-flash/llms.txt --- # Qwen3.7 Plus Model id on AIHubMix: `qwen3.7-plus` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > The Qwen 3.7 series' mid-to-high cost-performance "Plus" model builds on strong text capabilities with a comprehensive upgrade to vision-language abilities, while retaining full agent capabilities in coding, tool use, and productivity workflows. Its core features are multimodal interactive hybrid agent capabilities, able to perceive real-world scenes, read screens and operate GUIs, generate code based on visual references, and provide end-to-end navigation of mobile applications. - Developer: Qwen - Context window: 1,000,000 tokens - Input modalities: text, image, video - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Structured outputs, Prompt caching - Release date: 2026-05-26 - Pricing: $0.282/M input tokens, $1.128/M output tokens, $0.0564/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/messages` — Anthropic Messages (`x-api-key: $AIHUBMIX_API_KEY` + `anthropic-version: 2023-06-01`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen3.7-plus","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"qwen3.7-plus","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/qwen3.7-plus - Try in browser: https://aihubmix.com/playground?model=qwen3.7-plus - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/qwen3.7-plus.ce9245bd.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen3.7-plus/llms.txt --- # Step 3.7 Flash Model id on AIHubMix: `step-3.7-flash` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > step-3.7-flash is stepfun's flagship inference model, designed for high-complexity tasks that require deep reasoning and fast execution. It excels at decomposing multi-step problems, performing tool calls, and maintaining consistency across massive datasets. It is the preferred choice for complex workloads such as long-context agents, advanced software engineering, and end-to-end research automation. - Developer: StepFun - Context window: 256,000 tokens - Input modalities: text, image, video - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Release date: 2026-05-29 - Pricing: $0.22/M input tokens, $1.32/M output tokens, $0.044/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"step-3.7-flash","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"step-3.7-flash","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/step-3.7-flash - Try in browser: https://aihubmix.com/playground?model=step-3.7-flash - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/step-3.7-flash.435fa0fd.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/step-3.7-flash/llms.txt --- # Claude Opus 4.8 Thinking Model id on AIHubMix: `claude-opus-4-8-think` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > The claude-opus-4-8-think model has adaptive thinking mode pre-enabled; the default thinking intensity is "medium", and it can be invoked directly via the OpenAI unified API. The claude-opus-4-8 model does not display thinking content by default; you need to set the extra parameter "display": "summarized" to enable it. - Developer: Anthropic - Context window: 1,000,000 tokens - Input modalities: text, image, PDF - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Computer use, Memory tool, Structured outputs, Citations, Prompt caching - Release date: 2026-05-28 - Pricing: $5/M input tokens, $25/M output tokens, $0.5/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"claude-opus-4-8-think","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"claude-opus-4-8-think","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/claude-opus-4-8-think - Try in browser: https://aihubmix.com/playground?model=claude-opus-4-8-think - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/claude-opus-4-8-think.f0e4aac7.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/claude-opus-4-8-think/llms.txt --- # Nemotron 3 Nano Omni 30B A3B (reasoning) (free) Model id on AIHubMix: `nemotron-3-nano-omni-30b-a3b-reasoning-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Developed by Nvidia, NVIDIA Nemotron™ 3 Nano Omni is a 30B-A3B open multimodal model designed to function as a perception and context sub-agent in enterprise agent systems. Supporting a massive context length of 256,000 tokens, it accepts and processes inputs across text, images, and video. This model provides powerful multimodal comprehension tailored for complex enterprise workflows. - Developer: Nvidia - Context window: 262,144 tokens - Input modalities: text, image, video, audio - Capabilities: Thinking, Streaming, Tool calling, Structured outputs - Release date: 2026-04-28 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"nemotron-3-nano-omni-30b-a3b-reasoning-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"nemotron-3-nano-omni-30b-a3b-reasoning-free","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/nemotron-3-nano-omni-30b-a3b-reasoning-free - Try in browser: https://aihubmix.com/playground?model=nemotron-3-nano-omni-30b-a3b-reasoning-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/nemotron-3-nano-omni-30b-a3b-reasoning-free.103a5565.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/nemotron-3-nano-omni-30b-a3b-reasoning-free/llms.txt --- # Nemotron 3 Ultra 550B A55B (free) Model id on AIHubMix: `nemotron-3-ultra-550b-a55b-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > NVIDIA Nemotron 3 Ultra is an open frontier-reasoning and orchestration model featuring an extensive 1,000,000-token context length. Built on a hybrid Transformer-Mamba mixture-of-experts (MoE) architecture, it operates with 55B active parameters out of a total of 550B parameters. This advanced structure is designed to support sophisticated reasoning and complex task orchestration. - Developer: Nvidia - Context window: 1,048,576 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Structured outputs - Release date: 2026-06-04 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"nemotron-3-ultra-550b-a55b-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"nemotron-3-ultra-550b-a55b-free","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/nemotron-3-ultra-550b-a55b-free - Try in browser: https://aihubmix.com/playground?model=nemotron-3-ultra-550b-a55b-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/nemotron-3-ultra-550b-a55b-free.2272fd6b.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/nemotron-3-ultra-550b-a55b-free/llms.txt --- # Qwen3.7 Max Model id on AIHubMix: `qwen3.7-max` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > The Max model, the largest and most capable in the Qwen3.7 series, is currently offering its pure-text model capabilities for trial. Qwen3.7 is a new-generation flagship model designed for the agent era; its core strengths lie in the breadth and depth of its agent capabilities: it performs excellently in programming, office and productivity tasks, and long-term autonomous execution. - Developer: Qwen - Context window: 1,000,000 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Structured outputs, Prompt caching - Release date: 2026-05-20 - Pricing: $1.69/M input tokens, $5.07/M output tokens, $0.169/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen3.7-max","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"qwen3.7-max","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/qwen3.7-max - Try in browser: https://aihubmix.com/playground?model=qwen3.7-max - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/qwen3.7-max.d5ff7fdf.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen3.7-max/llms.txt --- # GPT Image 2 Model id on AIHubMix: `gpt-image-2` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-image-2 is OpenAI's latest cutting-edge image generation model. Key value adds include better performance, quality, editing controls, and face preservation. (Currently the model's image generation time may be >5 minutes; it is recommended to set the client timeout to ≥10 minutes.) The model supports high input_fidelity and adding/removing one aspect of the image while retaining others. This model includes improvements in aspect ratio, resolution, and editing capabilities. - Developer: OpenAI - Input modalities: text, image - Release date: 2026-04-21 - Pricing: per-generation — see https://aihubmix.com/model/gpt-image-2 ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/edits` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-image-2","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `image` | - | no | | | | `images` | array,null | no | | | | `mask` | string | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `output_format` | string,null | no | "png" | | | `prompt` | string | yes | | | | `size` | string|string|null | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/gpt-image-2/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/gpt-image-2/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/gpt-image-2 - Try in browser: https://aihubmix.com/playground?model=gpt-image-2 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-image-2/llms.txt --- # Nemotron 3.5 Content Safety (free) Model id on AIHubMix: `nemotron-3.5-content-safety-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Developed by NVIDIA, Nemotron 3.5 Content Safety is a compact 4B-parameter multimodal guardrail model fine-tuned from Google Gemma-3-4B. Supporting an expansive context length of 128,000 tokens, this model moderates both user inputs and generated responses for LLMs and VLMs. It provides robust safety filtering to ensure aligned and secure interactions across multiple modalities. - Developer: Nvidia - Context window: 131,072 tokens - Input modalities: text, image - Capabilities: Thinking - Release date: 2026-06-04 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"nemotron-3.5-content-safety-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"nemotron-3.5-content-safety-free","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/nemotron-3.5-content-safety-free - Try in browser: https://aihubmix.com/playground?model=nemotron-3.5-content-safety-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/nemotron-3.5-content-safety-free.7bbef780.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/nemotron-3.5-content-safety-free/llms.txt --- # Coding GLM 5.2 Model id on AIHubMix: `coding-glm-5.2` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Currently, the special resources for this model are limited, but due to its popularity, the usage is too high, which may result in a large number of 429 errors. Resources have been coordinated, and improvements are expected in the coming weeks, so please stay tuned. In the meantime, it is recommended to use the regular-priced API model, which can ensure absolute stability. The model ID is glm-5.2. - Developer: Z.AI - Context window: 1,000,000 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2026-06-16 - Pricing: $0.06/M input tokens, $0.22/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"coding-glm-5.2","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"coding-glm-5.2","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/coding-glm-5.2 - Try in browser: https://aihubmix.com/playground?model=coding-glm-5.2 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/coding-glm-5.2.bd80088e.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/coding-glm-5.2/llms.txt --- # ERNIE 5.1 Model id on AIHubMix: `ernie-5.1` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > ERNIE 5.1 is the latest model in the Wenxin series, with comprehensive upgrades to its foundational capabilities and significant improvements in agents, knowledge, reasoning, and deep search. This upgrade uses a decoupled fully-asynchronous reinforcement learning technique to specifically address challenges encountered as large models evolve toward agent-based autonomous decision-making, such as training–inference numerical bias, low utilization of heterogeneous resources, and global issues caused by long-tail effects. It is paired with scaled agent post-training techniques to enhance model capabilities and generalization, enabling a three-step collaboration of environment, expert, and fusion that both ensures training efficiency and significantly improves the model’s stability and performance on complex tasks. - Developer: Baidu - Context window: 119,000 tokens - Input modalities: text - Capabilities: Thinking, Tool calling, Web search - Release date: 2026-05-10 - Pricing: $0.5634/M input tokens, $2.5353/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"ernie-5.1","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"ernie-5.1","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/ernie-5.1 - Try in browser: https://aihubmix.com/playground?model=ernie-5.1 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/ernie-5.1.4f286e45.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/ernie-5.1/llms.txt --- # Gemini 3.1 Flash Lite Model id on AIHubMix: `gemini-3.1-flash-lite` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > gemini-3.1-flash-lite is currently Google's latest and most cost-effective model, optimized for large-scale agent-based tasks, translation, and simple data processing. - Developer: Google - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio, PDF - Capabilities: Thinking, Tool calling, Web search, URL context, Code interpreter, File search, Structured outputs, Prompt caching - Release date: 2026-05-07 - Pricing: $0.25/M input tokens, $1.5/M output tokens, $0.025/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /gemini/v1beta/models/gemini-3.1-flash-lite:generateContent` — Google Gemini (`x-goog-api-key: $AIHUBMIX_API_KEY`; streaming: `:streamGenerateContent?alt=sse`) - `POST /v1/messages` — Anthropic Messages (`x-api-key: $AIHUBMIX_API_KEY` + `anthropic-version: 2023-06-01`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-3.1-flash-lite","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gemini-3.1-flash-lite","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gemini-3.1-flash-lite - Try in browser: https://aihubmix.com/playground?model=gemini-3.1-flash-lite - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gemini-3.1-flash-lite.6ba37bd5.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-3.1-flash-lite/llms.txt --- # Gemini 3.1 Flash Lite (no think) Model id on AIHubMix: `gemini-3.1-flash-lite-nothink` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > gemini-3.1-flash-lite is currently Google's latest and most cost-effective model, optimized for large-scale agent-based tasks, translation, and simple data processing. - Developer: Google - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio, PDF - Capabilities: Thinking, Tool calling, Web search, URL context, Code interpreter, File search, Structured outputs, Prompt caching - Pricing: $0.25/M input tokens, $1.5/M output tokens, $0.025/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /gemini/v1beta/models/gemini-3.1-flash-lite-nothink:generateContent` — Google Gemini (`x-goog-api-key: $AIHUBMIX_API_KEY`; streaming: `:streamGenerateContent?alt=sse`) - `POST /v1/messages` — Anthropic Messages (`x-api-key: $AIHUBMIX_API_KEY` + `anthropic-version: 2023-06-01`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-3.1-flash-lite-nothink","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gemini-3.1-flash-lite-nothink","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gemini-3.1-flash-lite-nothink - Try in browser: https://aihubmix.com/playground?model=gemini-3.1-flash-lite-nothink - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gemini-3.1-flash-lite-nothink.08c6c38c.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-3.1-flash-lite-nothink/llms.txt --- # Grok 4.3 Model id on AIHubMix: `grok-4.3` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Grok 4.3 is amongst the leading models in intelligence and well priced when comparing to other models of similar price. It's also notably fast, however very verbose. The model supports text and image input, outputs text, and has a 1m tokens context window. - Developer: Grok - Context window: 1,000,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Release date: 2026-04-17 - Pricing: $1.25/M input tokens, $2.5/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"grok-4.3","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"grok-4.3","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/grok-4.3 - Try in browser: https://aihubmix.com/playground?model=grok-4.3 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/grok-4.3.add5843e.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/grok-4.3/llms.txt --- # Happyhorse 1.0 I2v Model id on AIHubMix: `happyhorse-1.0-i2v` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > HappyHorse-1.0-I2V supports image-to-video generation, featuring highly faithful dynamic visual generation capabilities, accurately understanding textual semantics and producing smooth, natural, and detail-rich high-quality videos. - Developer: Qwen - Input modalities: text, image - Pricing: per-generation — see https://aihubmix.com/model/happyhorse-1.0-i2v ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /ai/v1/videos/{id}` until a terminal state - `POST /v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /v1/videos/{id}` until a terminal state ## Example ```bash curl -s https://aihubmix.com/ai/v1/videos \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"happyhorse-1.0-i2v","prompt":""}' ``` The response contains a task id — poll `GET https://aihubmix.com/ai/v1/videos/` until `status` is terminal, then read the output URL(s). Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (video endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `duration` | integer|null | no | 5 | | | `extra` | object,null | no | | | | `frame_images` | array,null | no | | | | `generate_audio` | boolean,null | no | | | | `model` | string | yes | | | | `prompt` | string | yes | | | | `resolution` | string,null | no | "720p" | 720p, 1080p, null | | `seed` | integer,null | no | | | | `size` | string,null | no | | | | `webhook_url` | string,null | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/happyhorse-1.0-i2v/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/happyhorse-1.0-i2v/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/happyhorse-1.0-i2v - Try in browser: https://aihubmix.com/playground?model=happyhorse-1.0-i2v - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/happyhorse-1.0-i2v/llms.txt --- # Happyhorse 1.0 R2v Model id on AIHubMix: `happyhorse-1.0-r2v` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > HappyHorse-1.0-R2V supports reference-guided video generation, offering more stable subject and scene referencing, supports up to 9 reference images, and can accurately maintain creative intent to achieve stronger expressive capabilities. - Developer: Qwen - Input modalities: text, image - Pricing: per-generation — see https://aihubmix.com/model/happyhorse-1.0-r2v ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /ai/v1/videos/{id}` until a terminal state - `POST /v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /v1/videos/{id}` until a terminal state ## Example ```bash curl -s https://aihubmix.com/ai/v1/videos \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"happyhorse-1.0-r2v","prompt":""}' ``` The response contains a task id — poll `GET https://aihubmix.com/ai/v1/videos/` until `status` is terminal, then read the output URL(s). Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (video endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string,null | no | "16:9" | 16:9, 9:16, 1:1, 4:3, 3:4, null | | `duration` | integer|null | no | 5 | | | `extra` | object,null | no | | | | `frame_images` | array,null | no | | | | `generate_audio` | boolean,null | no | | | | `input_references` | array,null | no | | | | `model` | string | yes | | | | `prompt` | string | yes | | | | `resolution` | string,null | no | "720p" | 720p, 1080p, null | | `seed` | integer,null | no | | | | `size` | string,null | no | | | | `webhook_url` | string,null | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/happyhorse-1.0-r2v/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/happyhorse-1.0-r2v/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/happyhorse-1.0-r2v - Try in browser: https://aihubmix.com/playground?model=happyhorse-1.0-r2v - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/happyhorse-1.0-r2v/llms.txt --- # Happyhorse 1.0 T2v Model id on AIHubMix: `happyhorse-1.0-t2v` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > HappyHorse-1.0-T2V supports text-to-video generation, featuring highly faithful dynamic visual generation capabilities, accurately understanding textual semantics and producing smooth, natural, and detail-rich high-quality videos. - Developer: Qwen - Input modalities: text - Pricing: per-generation — see https://aihubmix.com/model/happyhorse-1.0-t2v ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /ai/v1/videos/{id}` until a terminal state - `POST /v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /v1/videos/{id}` until a terminal state ## Example ```bash curl -s https://aihubmix.com/ai/v1/videos \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"happyhorse-1.0-t2v","prompt":""}' ``` The response contains a task id — poll `GET https://aihubmix.com/ai/v1/videos/` until `status` is terminal, then read the output URL(s). Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (video endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string,null | no | "16:9" | 16:9, 9:16, 1:1, 4:3, 3:4, null | | `duration` | integer|null | no | 5 | | | `extra` | object,null | no | | | | `generate_audio` | boolean,null | no | | | | `model` | string | yes | | | | `prompt` | string | yes | | | | `resolution` | string,null | no | "720p" | 720p, 1080p, null | | `seed` | integer,null | no | | | | `size` | string,null | no | | | | `webhook_url` | string,null | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/happyhorse-1.0-t2v/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/happyhorse-1.0-t2v/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/happyhorse-1.0-t2v - Try in browser: https://aihubmix.com/playground?model=happyhorse-1.0-t2v - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/happyhorse-1.0-t2v/llms.txt --- # Happyhorse 1.0 Video Edit Model id on AIHubMix: `happyhorse-1.0-video-edit` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > HappyHorse-1.0-Video-Edit supports video editing, allows editing videos via natural language commands, can reference up to 5 images to edit video elements locally or globally, and can accurately replicate video dynamics to achieve stronger expressive capabilities. - Developer: Qwen - Input modalities: text, video, image - Pricing: per-generation — see https://aihubmix.com/model/happyhorse-1.0-video-edit ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /ai/v1/videos/{id}` until a terminal state - `POST /v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /v1/videos/{id}` until a terminal state ## Example ```bash curl -s https://aihubmix.com/ai/v1/videos \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"happyhorse-1.0-video-edit","prompt":""}' ``` The response contains a task id — poll `GET https://aihubmix.com/ai/v1/videos/` until `status` is terminal, then read the output URL(s). Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (video endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `extra` | object,null | no | | | | `frame_images` | array,null | no | | | | `generate_audio` | boolean,null | no | | | | `input_references` | array,null | no | | | | `model` | string | yes | | | | `prompt` | string | yes | | | | `resolution` | string,null | no | "720p" | 720p, 1080p, null | | `seed` | integer,null | no | | | | `size` | string,null | no | | | | `webhook_url` | string,null | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/happyhorse-1.0-video-edit/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/happyhorse-1.0-video-edit/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/happyhorse-1.0-video-edit - Try in browser: https://aihubmix.com/playground?model=happyhorse-1.0-video-edit - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/happyhorse-1.0-video-edit/llms.txt --- # North Mini Code (free) Model id on AIHubMix: `north-mini-code-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Developed by Cohere, north-mini-code-free is the debut model of the North family and Cohere's first agentic coding model. This sparse mixture-of-experts model features 30B total parameters and 3B active parameters, designed and optimized for high performance. With an expansive context length of 256,000 tokens, it is well-suited for handling complex developer workflows and large codebases. - Developer: Cohere - Context window: 256,000 tokens - Input modalities: text - Capabilities: Thinking, Tool calling, Structured outputs - Release date: 2026-06-09 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"north-mini-code-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"north-mini-code-free","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/north-mini-code-free - Try in browser: https://aihubmix.com/playground?model=north-mini-code-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/north-mini-code-free.63c2c2b8.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/north-mini-code-free/llms.txt --- # GPT 5.5 Model id on AIHubMix: `gpt-5.5` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-5.5 raises the baseline for complex production workflows. It’s a strong fit for coding use cases, tool-heavy agents, grounded assistants, long-context retrieval, product-spec-to-plan workflows, and customer-facing workflows where execution quality and response polish are critical. - Developer: OpenAI - Context window: 1,050,000 tokens - Input modalities: text, image, PDF - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Computer use, File search, Structured outputs, Prompt caching, Server-side sessions - Release date: 2026-04-24 - Pricing: $5/M input tokens, $30/M output tokens, $0.5/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/messages` — Anthropic Messages (`x-api-key: $AIHUBMIX_API_KEY` + `anthropic-version: 2023-06-01`) - `POST /v1/responses` — OpenAI Responses (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-5.5","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-5.5","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-5.5 - Try in browser: https://aihubmix.com/playground?model=gpt-5.5 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-5.5.3fd607ba.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-5.5/llms.txt --- # GPT 5.5 Pro Model id on AIHubMix: `gpt-5.5-pro` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Please note: this model is extremely expensive and very slow. If a request fails due to network issues, you may still be charged heavily; we cannot refund charges incurred by requests to this model. GPT-5.5 pro is available in the Responses API only to enable support for multi-turn model interactions before responding to API requests, and other advanced API features in the future. Since GPT-5.5 pro is designed to tackle tough problems, some requests may take several minutes to finish. To avoid timeout, please set a longer timeout duration. It is recommended to use this under good network conditions. - Developer: OpenAI - Context window: 1,050,000 tokens - Input modalities: text, image - Capabilities: Thinking, Tool calling, Web search, Code interpreter, File search, Structured outputs, Prompt caching, Background mode - Release date: 2026-04-23 - Pricing: $30/M input tokens, $180/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-5.5-pro","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-5.5-pro","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-5.5-pro - Try in browser: https://aihubmix.com/playground?model=gpt-5.5-pro - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-5.5-pro.ac9761c0.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-5.5-pro/llms.txt --- # DeepSeek V4 Flash Model id on AIHubMix: `deepseek-v4-flash` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > (This model currently points to the older 0423 version; if you need to request the latest version, you can choose the model deepseek-v4-flash-0731)DeepSeek-V4 features an ultra-long context of one million characters and achieves leading performance domestically and in the open-source domain in agent capabilities, world knowledge, and reasoning. - Developer: DeepSeek - Context window: 1,000,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2026-04-24 - Pricing: $0.142/M input tokens, $0.284/M output tokens, $0.0284/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/messages` — Anthropic Messages (`x-api-key: $AIHUBMIX_API_KEY` + `anthropic-version: 2023-06-01`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"deepseek-v4-flash","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"deepseek-v4-flash","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/deepseek-v4-flash - Try in browser: https://aihubmix.com/playground?model=deepseek-v4-flash - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/deepseek-v4-flash.39a1ae2c.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/deepseek-v4-flash/llms.txt --- # DeepSeek V4 Pro Model id on AIHubMix: `deepseek-v4-pro` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > (This model currently points to the older 0423 version; if you need to request the latest version, you can choose the model deepseek-v4-pro-0813)DeepSeek-V4 features an ultra-long context of one million characters and achieves leading performance domestically and in the open-source domain in agent capabilities, world knowledge, and reasoning. - Developer: DeepSeek - Context window: 1,000,000 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2026-04-24 - Pricing: $1.69/M input tokens, $3.38/M output tokens, $0.1403/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"deepseek-v4-pro","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"deepseek-v4-pro","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/deepseek-v4-pro - Try in browser: https://aihubmix.com/playground?model=deepseek-v4-pro - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/deepseek-v4-pro.6d924ba4.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/deepseek-v4-pro/llms.txt --- # Gemma 4 31B It (free) Model id on AIHubMix: `gemma-4-31b-it-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Gemma 4 31B Instruct is a 30.7B dense multimodal model developed by Google DeepMind that supports both text and image inputs with text outputs. It features an expansive 256K token context window alongside a configurable thinking and reasoning mode. The model also includes native function support to streamline complex tasks. - Developer: Google - Context window: 262,144 tokens - Input modalities: text, image, video - Capabilities: Thinking, Tool calling, Web search, Document input - Release date: 2026-04-02 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemma-4-31b-it-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gemma-4-31b-it-free","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gemma-4-31b-it-free - Try in browser: https://aihubmix.com/playground?model=gemma-4-31b-it-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gemma-4-31b-it-free.3a731fa4.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemma-4-31b-it-free/llms.txt --- # Command A Plus 05 2026 Model id on AIHubMix: `command-a-plus-05-2026` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Cohere's stronger command model for multilingual agents and enterprise workflows - Developer: Cohere - Context window: 128,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Structured outputs - Release date: 2026-05-20 - Pricing: $2.5/M input tokens, $10/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/responses` — OpenAI Responses (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"command-a-plus-05-2026","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"command-a-plus-05-2026","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/command-a-plus-05-2026 - Try in browser: https://aihubmix.com/playground?model=command-a-plus-05-2026 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/command-a-plus-05-2026.fced225a.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/command-a-plus-05-2026/llms.txt --- # ERNIE 5.0 Model id on AIHubMix: `ernie-5.0` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > ERNIE 5.0 is the next-generation natively multimodal foundation model in the ERNIE family. Built on a unified multimodal architecture, it jointly learns from text, images, audio, and video to deliver broad multimodal capabilities. ERNIE 5.0 features significantly upgraded core capabilities and shows strong performance across benchmarks, with notable gains in multimodal understanding, instruction following, creative writing, factual accuracy, and agent planning with tool use. - Developer: Baidu - Context window: 119,000 tokens - Input modalities: text, image - Capabilities: Thinking, Tool calling, Web search - Pricing: $0.8219/M input tokens, $3.2877/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"ernie-5.0","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"ernie-5.0","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/ernie-5.0 - Try in browser: https://aihubmix.com/playground?model=ernie-5.0 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/ernie-5.0.f0273a09.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/ernie-5.0/llms.txt --- # Kimi K2.6 Model id on AIHubMix: `kimi-k2.6` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Kimi K2.6 is Kimi's latest and most intelligent model, with stronger and more stable long-range code-writing capabilities, significantly improved instruction-following and self-correction abilities, and support for text, image, and video inputs, thinking and non-thinking modes, as well as dialogue and Agent tasks.The model has a context length of 256k, supports long-form thinking, and excels at deep reasoning. - Developer: Moonshot AI - Context window: 262,144 tokens - Input modalities: text, image, video - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Release date: 2026-04-21 - Pricing: $0.95/M input tokens, $3.9995/M output tokens, $0.1608/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"kimi-k2.6","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"kimi-k2.6","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/kimi-k2.6 - Try in browser: https://aihubmix.com/playground?model=kimi-k2.6 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/kimi-k2.6.cc97fa02.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/kimi-k2.6/llms.txt --- # Claude Opus 4.7 Model id on AIHubMix: `claude-opus-4-7` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Claude Opus 4.7 is Anthropic’s latest and most powerful publicly available model. It has high autonomy and performs exceptionally well on long-horizon agent tasks, knowledge work, vision tasks, and memory tasks. The claude-opus-4-7 model does not display thinking content by default; you need to set the extra parameter "display": "summarized" to enable it.This page summarizes all the new features at release. - Developer: Anthropic - Context window: 1,000,000 tokens - Input modalities: text, image, PDF - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Computer use, Memory tool, Structured outputs, Citations, Prompt caching - Release date: 2026-04-16 - Pricing: $5/M input tokens, $25/M output tokens, $0.5/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"claude-opus-4-7","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"claude-opus-4-7","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/claude-opus-4-7 - Try in browser: https://aihubmix.com/playground?model=claude-opus-4-7 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/claude-opus-4-7.a0fbd201.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/claude-opus-4-7/llms.txt --- # Claude Opus 4.7 Thinking Model id on AIHubMix: `claude-opus-4-7-think` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > The claude-opus-4-7-think model has adaptive thinking mode pre-enabled; the default thinking intensity is "medium", and it can be invoked directly via the OpenAI unified API. The claude-opus-4-7 model does not display thinking content by default; you need to set the extra parameter "display": "summarized" to enable it. - Developer: Anthropic - Context window: 1,000,000 tokens - Input modalities: text, image, PDF - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Computer use, Memory tool, Structured outputs, Citations, Prompt caching - Release date: 2026-04-16 - Pricing: $5/M input tokens, $25/M output tokens, $0.5/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"claude-opus-4-7-think","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"claude-opus-4-7-think","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/claude-opus-4-7-think - Try in browser: https://aihubmix.com/playground?model=claude-opus-4-7-think - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/claude-opus-4-7-think.5a788964.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/claude-opus-4-7-think/llms.txt --- # GPT Chat Model id on AIHubMix: `gpt-chat-latest` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT Chat Latest points to OpenAI's stable API alias chat-latest that always resolves to the latest Instant chat model used in ChatGPT. As OpenAI rolls out new Instant model updates in the future, they are routed behind this slug automatically. - Developer: OpenAI - Context window: 400,000 tokens - Input modalities: text, image - Capabilities: Streaming, Tool calling, Web search, Code interpreter, File search, Structured outputs, Prompt caching - Release date: 2026-05-05 - Pricing: $5/M input tokens, $30/M output tokens, $0.5/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-chat-latest","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-chat-latest","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-chat-latest - Try in browser: https://aihubmix.com/playground?model=gpt-chat-latest - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-chat-latest.74f687e8.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-chat-latest/llms.txt --- # Qwen3.6 27B Model id on AIHubMix: `qwen3.6-27b` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > The Qwen3.6 series 27B native vision-language Dense model. Compared with the 3.5-27B, the model notably improves Agentic coding capability and further enhances STEM and reasoning abilities; on the visual modality side, spatial intelligence, object localization and detection capabilities are significantly strengthened, and video understanding, document OCR, and visual agent capabilities have steadily improved. - Developer: Qwen - Context window: 262,144 tokens - Input modalities: text, image, video - Capabilities: Thinking, Tool calling, Web search, Structured outputs - Release date: 2026-04-22 - Pricing: $0.422/M input tokens, $2.532/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen3.6-27b","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"qwen3.6-27b","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/qwen3.6-27b - Try in browser: https://aihubmix.com/playground?model=qwen3.6-27b - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/qwen3.6-27b.0cf4be70.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen3.6-27b/llms.txt --- # Qwen3.6 35B A3B Model id on AIHubMix: `qwen3.6-35b-a3b` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Qwen 3.6, the native vision-language Plus series model, demonstrates outstanding performance comparable to the current top cutting-edge models, with a significant improvement over the 3.5 series. The model's capabilities have been markedly enhanced in Agentic coding, front-end programming, Vibe coding and other coding abilities, universal multimodal recognition, OCR, object localization, and more. - Developer: Qwen - Context window: 262,144 tokens - Input modalities: text, image, video - Capabilities: Thinking, Tool calling, Web search, Code interpreter, Structured outputs - Release date: 2026-04-16 - Pricing: $0.254/M input tokens, $1.524/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen3.6-35b-a3b","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"qwen3.6-35b-a3b","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/qwen3.6-35b-a3b - Try in browser: https://aihubmix.com/playground?model=qwen3.6-35b-a3b - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/qwen3.6-35b-a3b.50217a89.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen3.6-35b-a3b/llms.txt --- # Qwen3.6 Flash Model id on AIHubMix: `qwen3.6-flash` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Qwen 3.6, the native vision-language Plus series model, demonstrates outstanding performance comparable to the current top cutting-edge models, with a significant improvement over the 3.5 series. The model's capabilities have been markedly enhanced in Agentic coding, front-end programming, Vibe coding and other coding abilities, universal multimodal recognition, OCR, object localization, and more. - Developer: Qwen - Context window: 1,000,000 tokens - Input modalities: text, image, video - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Structured outputs, Prompt caching - Release date: 2026-04-27 - Pricing: $0.169/M input tokens, $1.014/M output tokens, $0.0169/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen3.6-flash","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"qwen3.6-flash","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/qwen3.6-flash - Try in browser: https://aihubmix.com/playground?model=qwen3.6-flash - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/qwen3.6-flash.ba32721d.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen3.6-flash/llms.txt --- # Gemma 4 26B A4B It (free) Model id on AIHubMix: `gemma-4-26b-a4b-it-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Gemma 4 26B A4B IT is an instruction-tuned Mixture-of-Experts (MoE) model developed by Google DeepMind. Featuring an expansive context length of 262,144 tokens, it delivers near-31B quality with highly efficient inference. Despite its 25.2B total parameters, only 3.8B are activated per token, making it an incredibly fast and cost-effective solution. - Developer: Google - Context window: 262,144 tokens - Input modalities: text, image, video - Capabilities: Thinking, Tool calling, Web search, Document input - Release date: 2026-04-02 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemma-4-26b-a4b-it-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gemma-4-26b-a4b-it-free","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gemma-4-26b-a4b-it-free - Try in browser: https://aihubmix.com/playground?model=gemma-4-26b-a4b-it-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gemma-4-26b-a4b-it-free.886392c9.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemma-4-26b-a4b-it-free/llms.txt --- # grok-4-20-non-reasoning Model id on AIHubMix: `grok-4-20-non-reasoning` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Grok 4.2 is xAI’s latest large language model, built for strong reasoning, multimodal understanding, and enterprise use. It improves instruction following, honesty, and calibration over earlier Grok versions, while supporting both single‑agent and multi‑agent workflows. Designed as a general‑purpose, truth‑seeking assistant, Grok 4.2 is well suited for research, analysis, coding, and complex professional tasks when deployed with appropriate guardrails. - Developer: Grok - Context window: 1,000,000 tokens - Input modalities: text, image - Capabilities: Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2026-03-09 - Pricing: $2/M input tokens, $6/M output tokens, $1/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"grok-4-20-non-reasoning","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"grok-4-20-non-reasoning","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/grok-4-20-non-reasoning - Try in browser: https://aihubmix.com/playground?model=grok-4-20-non-reasoning - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/grok-4-20-non-reasoning.b07a3db5.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/grok-4-20-non-reasoning/llms.txt --- # Grok 4 20 (reasoning) Model id on AIHubMix: `grok-4-20-reasoning` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Grok 4.2 is xAI’s latest large language model, built for strong reasoning, multimodal understanding, and enterprise use. It improves instruction following, honesty, and calibration over earlier Grok versions, while supporting both single‑agent and multi‑agent workflows. Designed as a general‑purpose, truth‑seeking assistant, Grok 4.2 is well suited for research, analysis, coding, and complex professional tasks when deployed with appropriate guardrails. - Developer: Grok - Context window: 1,000,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2026-03-09 - Pricing: $2/M input tokens, $6/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"grok-4-20-reasoning","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"grok-4-20-reasoning","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/grok-4-20-reasoning - Try in browser: https://aihubmix.com/playground?model=grok-4-20-reasoning - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/grok-4-20-reasoning.9951083f.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/grok-4-20-reasoning/llms.txt --- # Qwen Image 2.0 Model id on AIHubMix: `qwen-image-2.0` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > The Qwen-Image-2.0 series accelerated models integrate image generation and image editing; they offer more professional text rendering with support for 1k-token instructions, finer realistic textures and delicate portrayal of photorealistic scenes, and stronger semantic adherence. The accelerated version effectively achieves an optimal balance between model quality and performance. - Developer: Qwen - Input modalities: text, image - Release date: 2026-03-03 - Pricing: per-generation — see https://aihubmix.com/model/qwen-image-2.0 ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/edits` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen-image-2.0","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `image` | - | no | | | | `images` | array,null | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `negative_prompt` | string,null | no | | | | `output_format` | string,null | no | "png" | png, jpeg, webp, null | | `prompt` | string | yes | | | | `response_format` | string,null | no | "url" | url, b64_json, null | | `seed` | integer,null | no | | | | `size` | string|null | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/qwen-image-2.0/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/qwen-image-2.0/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/qwen-image-2.0 - Try in browser: https://aihubmix.com/playground?model=qwen-image-2.0 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen-image-2.0/llms.txt --- # Qwen Image 2.0 Pro Model id on AIHubMix: `qwen-image-2.0-pro` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > The Qwen-Image-2.0 full-powered models achieve the integration of image generation and image editing; they offer more professional text rendering with support for 1k-token instructions, more refined photorealistic textures and delicate depiction of realistic scenes, and stronger semantic adherence. The full-powered version delivers the strongest text rendering capability and realism in the 2.0 series. - Developer: Qwen - Input modalities: text, image - Release date: 2026-03-03 - Pricing: per-generation — see https://aihubmix.com/model/qwen-image-2.0-pro ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/edits` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen-image-2.0-pro","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `image` | - | no | | | | `images` | array,null | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `negative_prompt` | string,null | no | | | | `output_format` | string,null | no | "png" | png, jpeg, webp, null | | `prompt` | string | yes | | | | `response_format` | string,null | no | "url" | url, b64_json, null | | `seed` | integer,null | no | | | | `size` | string|null | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/qwen-image-2.0-pro/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/qwen-image-2.0-pro/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/qwen-image-2.0-pro - Try in browser: https://aihubmix.com/playground?model=qwen-image-2.0-pro - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen-image-2.0-pro/llms.txt --- # Coding MiniMax M3 (free) Model id on AIHubMix: `coding-minimax-m3-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > coding-minimax-m3-free is a free and open version offered by AIHubMix specifically for MiniMax users. To maintain reliable service, each account is limited to 5 requests per minute, 500 requests per day, and 1 million tokens per day. - Developer: Minimax - Context window: 1,000,000 tokens - Input modalities: text, image, video - Capabilities: Thinking, Streaming, Tool calling, Web search, Prompt caching - Release date: 2026-05-31 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"coding-minimax-m3-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"coding-minimax-m3-free","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/coding-minimax-m3-free - Try in browser: https://aihubmix.com/playground?model=coding-minimax-m3-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/coding-minimax-m3-free.dc23cde9.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/coding-minimax-m3-free/llms.txt --- # Doubao Seedance 2.0 260128 Model id on AIHubMix: `doubao-seedance-2-0-260128` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > The Doubao large-model team has launched a new-generation professional-grade multimodal video-creation model, Seedance 2.0. It supports images, videos, and audio as multimodal reference inputs to generate videos, and also offers video editing and extension capabilities. It can accurately reproduce various details while maintaining stable character features, delivers highly realistic audiovisual stability, and is deeply adapted to core scenarios such as commercial advertising, film and TV production, and social media marketing. - Developer: ByteDance - Input modalities: text, image, video, audio - Release date: 2026-01-28 - Pricing: per-generation — see https://aihubmix.com/model/doubao-seedance-2-0-260128 ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /ai/v1/videos/{id}` until a terminal state - `POST /v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /v1/videos/{id}` until a terminal state ## Example ```bash curl -s https://aihubmix.com/ai/v1/videos \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"doubao-seedance-2-0-260128","prompt":""}' ``` The response contains a task id — poll `GET https://aihubmix.com/ai/v1/videos/` until `status` is terminal, then read the output URL(s). Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (video endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string|string|null | no | "adaptive" | | | `duration` | integer|null | no | 5 | | | `extra` | object,null | no | | | | `frame_images` | array,null | no | | | | `generate_audio` | boolean,null | no | true | | | `input_references` | array,null | no | | | | `model` | string | yes | | | | `prompt` | string | yes | | | | `resolution` | string|null | no | "720p" | | | `webhook_url` | string,null | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/doubao-seedance-2-0-260128/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/doubao-seedance-2-0-260128/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/doubao-seedance-2-0-260128 - Try in browser: https://aihubmix.com/playground?model=doubao-seedance-2-0-260128 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/doubao-seedance-2-0-260128/llms.txt --- # Doubao Seedance 2.0 Fast 260128 Model id on AIHubMix: `doubao-seedance-2-0-fast-260128` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Seedance 2.0 fast is a next-generation multimodal video-creation model launched by the Doubao large-model team; it inherits the core features and advantages of Seedance 2.0 and generates content faster. - Developer: Doubao - Input modalities: text, image, video, audio - Release date: 2026-01-28 - Pricing: per-generation — see https://aihubmix.com/model/doubao-seedance-2-0-fast-260128 ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /ai/v1/videos/{id}` until a terminal state - `POST /v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /v1/videos/{id}` until a terminal state ## Example ```bash curl -s https://aihubmix.com/ai/v1/videos \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"doubao-seedance-2-0-fast-260128","prompt":""}' ``` The response contains a task id — poll `GET https://aihubmix.com/ai/v1/videos/` until `status` is terminal, then read the output URL(s). Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (video endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string|string|null | no | "adaptive" | | | `duration` | integer|null | no | 5 | | | `extra` | object,null | no | | | | `frame_images` | array,null | no | | | | `generate_audio` | boolean,null | no | true | | | `input_references` | array,null | no | | | | `model` | string | yes | | | | `prompt` | string | yes | | | | `resolution` | string|null | no | "720p" | | | `webhook_url` | string,null | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/doubao-seedance-2-0-fast-260128/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/doubao-seedance-2-0-fast-260128/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/doubao-seedance-2-0-fast-260128 - Try in browser: https://aihubmix.com/playground?model=doubao-seedance-2-0-fast-260128 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/doubao-seedance-2-0-fast-260128/llms.txt --- # Doubao Seedance 2.0 Mini 260615 Model id on AIHubMix: `doubao-seedance-2-0-mini-260615` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Seedance 2.0 mini is a next-generation, cost-effective video generation model launched to address broader video-generation needs. While maintaining competitive performance, it brings video-generation capabilities to lower barriers to entry, more frequent use, and more scalable/large-scale application scenarios. - Developer: ByteDance - Input modalities: text, image, video, audio - Release date: 2026-06-15 - Pricing: per-generation — see https://aihubmix.com/model/doubao-seedance-2-0-mini-260615 ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /ai/v1/videos/{id}` until a terminal state - `POST /v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /v1/videos/{id}` until a terminal state ## Example ```bash curl -s https://aihubmix.com/ai/v1/videos \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"doubao-seedance-2-0-mini-260615","prompt":""}' ``` The response contains a task id — poll `GET https://aihubmix.com/ai/v1/videos/` until `status` is terminal, then read the output URL(s). Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (video endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string|string|null | no | "adaptive" | | | `duration` | integer|null | no | 5 | | | `extra` | object,null | no | | | | `frame_images` | array,null | no | | | | `generate_audio` | boolean,null | no | true | | | `input_references` | array,null | no | | | | `model` | string | yes | | | | `prompt` | string | yes | | | | `resolution` | string|null | no | "720p" | | | `webhook_url` | string,null | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/doubao-seedance-2-0-mini-260615/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/doubao-seedance-2-0-mini-260615/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/doubao-seedance-2-0-mini-260615 - Try in browser: https://aihubmix.com/playground?model=doubao-seedance-2-0-mini-260615 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/doubao-seedance-2-0-mini-260615/llms.txt --- # GLM 5.1 Model id on AIHubMix: `glm-5.1` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GLM-5.1 is Zhipu's latest flagship model, with greatly enhanced coding capabilities and significantly improved long-range task performance. It can continuously and autonomously work for up to 8 hours on a single task, completing the full closed loop from planning and execution to iterative optimization, delivering engineering-grade results. In terms of general capability and coding ability, GLM-5.1's overall performance aligns with Claude Opus 4.6, and it demonstrates stronger sustained work capability in long-range autonomous execution, complex engineering optimization, and real-world development scenarios, making it an ideal foundation for building Autonomous Agents and long-horizon Coding Agents. - Developer: Z.AI - Context window: 200,000 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2026-04-07 - Pricing: $0.845/M input tokens, $3.38/M output tokens, $0.1831/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/messages` — Anthropic Messages (`x-api-key: $AIHUBMIX_API_KEY` + `anthropic-version: 2023-06-01`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"glm-5.1","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"glm-5.1","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/glm-5.1 - Try in browser: https://aihubmix.com/playground?model=glm-5.1 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/glm-5.1.32dd5058.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/glm-5.1/llms.txt --- # GLM Image Model id on AIHubMix: `glm-image` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GLM-Image is Zhipu AI's new flagship image generation model. The model is trained entirely on domestic chips and adopts an original hybrid architecture combining "autoregressive + diffusion decoder," balancing global instruction understanding with local detail depiction. It overcomes generation challenges in knowledge-intensive scenarios such as posters, PPTs, and popular science illustrations. This represents an important exploration towards the new generation of "cognitive generation" technology paradigms exemplified by Nano Banana Pro. - Developer: Z.AI - Input modalities: text - Pricing: per-generation — see https://aihubmix.com/model/glm-image ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"glm-image","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `prompt` | string | yes | | | | `size` | string|null | no | "1280x1280" | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/glm-image/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/glm-image/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/glm-image - Try in browser: https://aihubmix.com/playground?model=glm-image - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/glm-image/llms.txt --- # Qwen3.6 Plus Model id on AIHubMix: `qwen3.6-plus` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Qwen 3.6, the native vision-language Plus series model, demonstrates outstanding performance comparable to the current top cutting-edge models, with a significant improvement over the 3.5 series. The model's capabilities have been markedly enhanced in Agentic coding, front-end programming, Vibe coding and other coding abilities, universal multimodal recognition, OCR, object localization, and more. - Developer: Qwen - Context window: 1,000,000 tokens - Input modalities: text, image, video - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Structured outputs, Prompt caching - Release date: 2026-04-02 - Pricing: $0.282/M input tokens, $1.692/M output tokens, $0.0282/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen3.6-plus","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"qwen3.6-plus","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/qwen3.6-plus - Try in browser: https://aihubmix.com/playground?model=qwen3.6-plus - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/qwen3.6-plus.03dea1da.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen3.6-plus/llms.txt --- # Wan2.7 I2v Model id on AIHubMix: `wan2.7-i2v` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Wanxiang 2.7 — image-to-video: performance capabilities comprehensively upgraded. Dramatic/dialogue scenes convey delicate, natural emotions; action scenes are fierce, with every punch landing hard. Combined with more dramatic, rhythmically paced camera cuts, it achieves stronger overall performance. - Developer: Qwen - Input modalities: text, image, audio - Pricing: per-generation — see https://aihubmix.com/model/wan2.7-i2v ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /ai/v1/videos/{id}` until a terminal state - `POST /v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /v1/videos/{id}` until a terminal state ## Example ```bash curl -s https://aihubmix.com/ai/v1/videos \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"wan2.7-i2v","prompt":""}' ``` The response contains a task id — poll `GET https://aihubmix.com/ai/v1/videos/` until `status` is terminal, then read the output URL(s). Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (video endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `duration` | integer|null | no | 5 | | | `extra` | object,null | no | | | | `frame_images` | array,null | no | | | | `generate_audio` | boolean,null | no | | | | `model` | string | yes | | | | `prompt` | string | yes | | | | `resolution` | string,null | no | "1080p" | 720p, 1080p, null | | `seed` | integer,null | no | | | | `webhook_url` | string,null | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/wan2.7-i2v/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/wan2.7-i2v/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/wan2.7-i2v - Try in browser: https://aihubmix.com/playground?model=wan2.7-i2v - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/wan2.7-i2v/llms.txt --- # Wan2.7 R2v Model id on AIHubMix: `wan2.7-r2v` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Wanxiang 2.7 — reference-driven video generation: more stable references for characters, props, and scenes; supports up to five mixed image/video references; supports audio/timbre references; combined with foundational capability upgrades to deliver stronger performance. - Developer: Qwen - Input modalities: text, image, video, audio - Pricing: per-generation — see https://aihubmix.com/model/wan2.7-r2v ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /ai/v1/videos/{id}` until a terminal state - `POST /v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /v1/videos/{id}` until a terminal state ## Example ```bash curl -s https://aihubmix.com/ai/v1/videos \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"wan2.7-r2v","prompt":""}' ``` The response contains a task id — poll `GET https://aihubmix.com/ai/v1/videos/` until `status` is terminal, then read the output URL(s). Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (video endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string|null | no | "16:9" | | | `duration` | integer|null | no | 5 | | | `extra` | object,null | no | | | | `frame_images` | array,null | no | | | | `generate_audio` | boolean,null | no | | | | `input_references` | array,null | no | | | | `model` | string | yes | | | | `prompt` | string | yes | | | | `resolution` | string|null | no | "1080p" | | | `seed` | integer,null | no | | | | `webhook_url` | string,null | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/wan2.7-r2v/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/wan2.7-r2v/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/wan2.7-r2v - Try in browser: https://aihubmix.com/playground?model=wan2.7-r2v - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/wan2.7-r2v/llms.txt --- # Wan2.7 T2v Model id on AIHubMix: `wan2.7-t2v` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Wanxiang 2.7 — text-to-video: performance capabilities comprehensively upgraded. Dramatic/dialogue scenes convey delicate, natural emotions; action scenes are intense, with every punch landing hard. Paired with more dramatic and rhythmically paced camera cuts, it achieves stronger overall acting performance. - Developer: Qwen - Input modalities: text, audio - Pricing: per-generation — see https://aihubmix.com/model/wan2.7-t2v ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /ai/v1/videos/{id}` until a terminal state - `POST /v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /v1/videos/{id}` until a terminal state ## Example ```bash curl -s https://aihubmix.com/ai/v1/videos \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"wan2.7-t2v","prompt":""}' ``` The response contains a task id — poll `GET https://aihubmix.com/ai/v1/videos/` until `status` is terminal, then read the output URL(s). Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (video endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string,null | no | "16:9" | 16:9, 9:16, 1:1, 4:3, 3:4, null | | `duration` | integer|null | no | 5 | | | `extra` | object,null | no | | | | `generate_audio` | boolean,null | no | | | | `model` | string | yes | | | | `prompt` | string | yes | | | | `resolution` | string,null | no | "1080p" | 720p, 1080p, null | | `seed` | integer,null | no | | | | `webhook_url` | string,null | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/wan2.7-t2v/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/wan2.7-t2v/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/wan2.7-t2v - Try in browser: https://aihubmix.com/playground?model=wan2.7-t2v - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/wan2.7-t2v/llms.txt --- # Wan2.7 Videoedit Model id on AIHubMix: `wan2.7-videoedit` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Wanxiang 2.7 — video editing: edit videos using natural-language commands, supporting local or global edits; can use reference images to replace video elements; supports replicating video actions, special effects, camera movements, and other dynamic processes. - Developer: Qwen - Input modalities: text, image, video - Pricing: per-generation — see https://aihubmix.com/model/wan2.7-videoedit ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /ai/v1/videos/{id}` until a terminal state - `POST /v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /v1/videos/{id}` until a terminal state ## Example ```bash curl -s https://aihubmix.com/ai/v1/videos \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"wan2.7-videoedit","prompt":""}' ``` The response contains a task id — poll `GET https://aihubmix.com/ai/v1/videos/` until `status` is terminal, then read the output URL(s). Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (video endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string,null | no | "16:9" | 16:9, 9:16, 1:1, 4:3, 3:4, null | | `duration` | integer|null | no | | | | `extra` | object,null | no | | | | `input_references` | array,null | no | | | | `model` | string | yes | | | | `prompt` | string | yes | | | | `resolution` | string,null | no | "1080p" | 720p, 1080p, null | | `seed` | integer,null | no | | | | `webhook_url` | string,null | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/wan2.7-videoedit/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/wan2.7-videoedit/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/wan2.7-videoedit - Try in browser: https://aihubmix.com/playground?model=wan2.7-videoedit - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/wan2.7-videoedit/llms.txt --- # CC K2.6 Code Preview Model id on AIHubMix: `cc-k2.6-code-preview` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > for claude code - Developer: Moonshot AI - Context window: 262,144 tokens - Input modalities: text, image, video - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Pricing: $0.2/M input tokens, $0.2/M output tokens, $0.02/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"cc-k2.6-code-preview","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"cc-k2.6-code-preview","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/cc-k2.6-code-preview - Try in browser: https://aihubmix.com/playground?model=cc-k2.6-code-preview - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/cc-k2.6-code-preview.ca91e626.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/cc-k2.6-code-preview/llms.txt --- # Gemma 4 26B A4B It Model id on AIHubMix: `gemma-4-26b-a4b-it` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > A Mixture-of-Experts model that activates only 4B parameters per inference,delivering high-performance reasoning with a fraction of the memory cost - idealfor cost-efficient, high-throughput server deployments. - Developer: Google - Context window: 262,144 tokens - Input modalities: text, image, video - Capabilities: Thinking, Tool calling, Web search, Document input - Release date: 2026-04-02 - Pricing: $0.14/M input tokens, $0.4/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemma-4-26b-a4b-it","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gemma-4-26b-a4b-it","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gemma-4-26b-a4b-it - Try in browser: https://aihubmix.com/playground?model=gemma-4-26b-a4b-it - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gemma-4-26b-a4b-it.5afdda31.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemma-4-26b-a4b-it/llms.txt --- # Gemma 4 31B It Model id on AIHubMix: `gemma-4-31b-it` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Gemma 4 31B Instruct is Google DeepMind's 30.7B dense multimodal model supporting text and image input with text output. Features a 256K token context window, configurable thinking/reasoning mode, native function calling, and multilingual support across 140+ languages. Strong on coding, reasoning, and document understanding tasks. Apache 2.0 license. - Developer: Google - Context window: 262,144 tokens - Input modalities: text, image, video - Capabilities: Thinking, Tool calling, Web search, Document input - Release date: 2026-04-02 - Pricing: $0.2/M input tokens, $0.4/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemma-4-31b-it","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gemma-4-31b-it","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gemma-4-31b-it - Try in browser: https://aihubmix.com/playground?model=gemma-4-31b-it - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gemma-4-31b-it.f3b636be.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemma-4-31b-it/llms.txt --- # GPT 5.4 Model id on AIHubMix: `gpt-5.4` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-5.4 is our frontier model for complex professional work.Reasoning.effort supports: none (default), low, medium, high and xhigh. - Developer: OpenAI - Context window: 1,050,000 tokens - Input modalities: text, image, PDF - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Computer use, File search, Structured outputs, Prompt caching, Server-side sessions - Release date: 2026-03-05 - Pricing: $2.5/M input tokens, $15/M output tokens, $0.25/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/responses` — OpenAI Responses (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/messages` — Anthropic Messages (`x-api-key: $AIHUBMIX_API_KEY` + `anthropic-version: 2023-06-01`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-5.4","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-5.4","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-5.4 - Try in browser: https://aihubmix.com/playground?model=gpt-5.4 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-5.4.ab6f054c.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-5.4/llms.txt --- # Wan2.7 Image Model id on AIHubMix: `wan2.7-image` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Wanxiang 2.7 — image generation and editing: supports text-to-image, text-to-multi-image, image-to-multi-image, image editing, multi-image reference generation, and interactive editing, with stronger performance in text rendering, subject consistency, and following complex instructions. - Developer: Qwen - Input modalities: text, image - Release date: 2026-05-29 - Pricing: per-generation — see https://aihubmix.com/model/wan2.7-image ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/edits` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"wan2.7-image","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string|null | no | | | | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `image` | - | no | | | | `images` | array,null | no | | | | `model` | string | yes | | | | `n` | integer,null | no | null | | | `output_format` | string,null | no | "png" | png, jpeg, webp, null | | `prompt` | string | yes | | | | `response_format` | string,null | no | "url" | url, b64_json, null | | `seed` | integer,null | no | | | | `size` | string|string|null | no | null | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/wan2.7-image/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/wan2.7-image/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/wan2.7-image - Try in browser: https://aihubmix.com/playground?model=wan2.7-image - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/wan2.7-image/llms.txt --- # Wan2.7 Image Pro Model id on AIHubMix: `wan2.7-image-pro` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Wanxiang 2.7 — image generation and editing: supports text-to-image, text-to-multi-image, image-to-multi-image, image editing, multi-image reference generation, and interactive editing, with stronger performance in text rendering, subject consistency, and following complex instructions. - Developer: Qwen - Input modalities: text, image - Release date: 2026-05-29 - Pricing: per-generation — see https://aihubmix.com/model/wan2.7-image-pro ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/edits` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"wan2.7-image-pro","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string|null | no | | | | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `image` | - | no | | | | `images` | array,null | no | | | | `model` | string | yes | | | | `n` | integer,null | no | null | | | `output_format` | string,null | no | "png" | png, jpeg, webp, null | | `prompt` | string | yes | | | | `response_format` | string,null | no | "url" | url, b64_json, null | | `seed` | integer,null | no | | | | `size` | string|string|null | no | null | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/wan2.7-image-pro/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/wan2.7-image-pro/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/wan2.7-image-pro - Try in browser: https://aihubmix.com/playground?model=wan2.7-image-pro - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/wan2.7-image-pro/llms.txt --- # Claude Sonnet 4.6 Model id on AIHubMix: `claude-sonnet-4-6` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Claude Sonnet 4.6 delivers frontier intelligence at scale—built for coding, agents, and enterprise workflows. Use cases include: Agents: Sonnet 4.6 excels at complex, multi-step tasks requiring sustained reasoning and adaptive decision-making—ideal for workflows where reliability and autonomy matter most. Coding: Sonnet 4.6 is built for iterative development work, handling complex codebases without losing quality as you guide it through building, refactoring, and debugging. It can compress multi-day projects into hours with the technical depth to deliver production-ready solutions. Enterprise workflows: Sonnet 4.6 powers agents that manage professional projects from start to finish, leveraging memory to maintain context across files with a step-change improvement in creating spreadsheets, slides, and docs. Financial analysis: Sonnet 4.6 connects the dots across regulatory filings, market reports, and internal data—enabling sophisticated modeling and proactive compliance. Cybersecurity: Sonnet 4.6 brings professional-grade analysis to security workflows, correlating logs, vulnerability databases, and threat intelligence for proactive threat detection and automated incident response. Computer use: Sonnet 4.6 delivers confident, consistent navigation with more human-like browsing—enabling better web QA, workflow automation, and advanced user experiences. - Developer: Anthropic - Context window: 1,000,000 tokens - Input modalities: text, image, PDF - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Computer use, Memory tool, Structured outputs, Citations, Prompt caching - Release date: 2026-02-17 - Pricing: $3/M input tokens, $15/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /gemini/v1beta/models/claude-sonnet-4-6:generateContent` — Google Gemini (`x-goog-api-key: $AIHUBMIX_API_KEY`; streaming: `:streamGenerateContent?alt=sse`) - `POST /v1/messages` — Anthropic Messages (`x-api-key: $AIHUBMIX_API_KEY` + `anthropic-version: 2023-06-01`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"claude-sonnet-4-6","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"claude-sonnet-4-6","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/claude-sonnet-4-6 - Try in browser: https://aihubmix.com/playground?model=claude-sonnet-4-6 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/claude-sonnet-4-6.3c713629.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/claude-sonnet-4-6/llms.txt --- # Doubao Seed 2.0 Lite 260428 Model id on AIHubMix: `doubao-seed-2-0-lite-260428` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Doubao Coding model optimized for real-world programming environments that can reliably invoke tools in common IDEs such as Claude Code. The model is specially optimized for frontend capabilities and performs well with common frontend frameworks. The model supports Skills and can work with various custom skills. - Developer: ByteDance - Context window: 256,000 tokens - Input modalities: text, image, video, document, audio - Capabilities: Thinking, Tool calling, Web search, File search, Document input, Structured outputs, Prompt caching, Server-side sessions - Release date: 2026-04-28 - Pricing: $0.0904/M input tokens, $0.5425/M output tokens, $0.0181/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"doubao-seed-2-0-lite-260428","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"doubao-seed-2-0-lite-260428","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/doubao-seed-2-0-lite-260428 - Try in browser: https://aihubmix.com/playground?model=doubao-seed-2-0-lite-260428 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/doubao-seed-2-0-lite-260428.3d215d6c.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/doubao-seed-2-0-lite-260428/llms.txt --- # Doubao Seed 2.0 Mini 260428 Model id on AIHubMix: `doubao-seed-2-0-mini-260428` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Doubao Coding model optimized for real-world programming environments that can reliably invoke tools in common IDEs such as Claude Code. The model is specially optimized for frontend capabilities and performs well with common frontend frameworks. The model supports Skills and can work with various custom skills. - Developer: ByteDance - Context window: 256,000 tokens - Input modalities: text, image, video, document, audio - Capabilities: Thinking, Tool calling, Web search, File search, Document input, Structured outputs, Prompt caching, Server-side sessions - Release date: 2026-04-28 - Pricing: $0.0301/M input tokens, $0.3014/M output tokens, $0.006/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"doubao-seed-2-0-mini-260428","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"doubao-seed-2-0-mini-260428","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/doubao-seed-2-0-mini-260428 - Try in browser: https://aihubmix.com/playground?model=doubao-seed-2-0-mini-260428 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/doubao-seed-2-0-mini-260428.b41769cb.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/doubao-seed-2-0-mini-260428/llms.txt --- # Gemini 3.1 Flash Image Preview Model id on AIHubMix: `gemini-3.1-flash-image-preview` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model is scheduled for retirement.** Calls still work until then. Plan to migrate to `gemini-3.1-flash-image`. Its guide: https://aihubmix.com/model/gemini-3.1-flash-image/llms.txt > > Once retired, requests return HTTP 404 `model_retired`. Responses already carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > gemini-3.1-flash-image-preview (Nano Banana 2) features professional-grade visual intelligence, lightning-fast efficiency, and realistic, grounded generative capabilities. This model serves as the high-efficiency counterpart to Gemini 3 Pro Image, optimized for speed and high-volume developer use cases. - Developer: Google - Input modalities: text, image - Release date: 2026-02-26 - Pricing: per-generation — see https://aihubmix.com/model/gemini-3.1-flash-image-preview ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/edits` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-3.1-flash-image-preview"}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string|string|null | no | | 1:1, 1:4, 4:1, 1:8, 8:1, 2:3, … | | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `image` | - | no | | | | `images` | array,null | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `negative_prompt` | string,null | no | | | | `output_format` | string,null | no | "png" | png, jpeg, webp, null | | `prompt` | string | no | | | | `reasoning_effort` | string | no | | minimal, high | | `response_format` | string,null | no | "url" | url, b64_json, null | | `seed` | integer,null | no | | | | `size` | string|string|string|null | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/gemini-3.1-flash-image-preview/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/gemini-3.1-flash-image-preview/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/gemini-3.1-flash-image-preview - Try in browser: https://aihubmix.com/playground?model=gemini-3.1-flash-image-preview - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-3.1-flash-image-preview/llms.txt --- # Gemini 3.1 Pro Preview Model id on AIHubMix: `gemini-3.1-pro-preview` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Gemini 3.1 Pro Preview is designed to further optimize the performance and reliability of the Gemini 3 Pro series, offering improved reasoning capabilities, greater token efficiency, and a more robust, factually consistent user experience. It is optimized for software-engineering behaviors and usability, and is also suitable for agent workflows that require precise tool invocation and reliable multi-step execution, enabling stable operation across a variety of real-world scenarios. - Developer: Google - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio, PDF - Capabilities: Thinking, Tool calling, Web search, URL context, Code interpreter, File search, Structured outputs, Prompt caching - Release date: 2026-02-19 - Pricing: $2/M input tokens, $12/M output tokens, $0.5/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /gemini/v1beta/models/gemini-3.1-pro-preview:generateContent` — Google Gemini (`x-goog-api-key: $AIHUBMIX_API_KEY`; streaming: `:streamGenerateContent?alt=sse`) - `POST /v1/messages` — Anthropic Messages (`x-api-key: $AIHUBMIX_API_KEY` + `anthropic-version: 2023-06-01`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-3.1-pro-preview","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gemini-3.1-pro-preview","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gemini-3.1-pro-preview - Try in browser: https://aihubmix.com/playground?model=gemini-3.1-pro-preview - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gemini-3.1-pro-preview.8ba876eb.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-3.1-pro-preview/llms.txt --- # Gemini 3.1 Pro Preview Customtools Model id on AIHubMix: `gemini-3.1-pro-preview-customtools` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > gemini-3.1-pro-preview-customtools For users who build applications mixing bash and custom tools, the Gemini 3.1 Pro preview provides a separate endpoint accessible via the API call gemini-3.1-pro-preview-customtools. This endpoint is better at prioritizing your custom tools (for example, view_file or search_code). Please note that while gemini-3.1-pro-preview-customtools is optimized for agent workflows that use custom tools and Bash, you may experience quality fluctuations in some use cases that cannot benefit from these tools. - Developer: Google - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio, PDF - Capabilities: Thinking, Tool calling, Web search, URL context, Code interpreter, File search, Structured outputs, Prompt caching - Release date: 2026-02-19 - Pricing: $2/M input tokens, $12/M output tokens, $0.5/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /gemini/v1beta/models/gemini-3.1-pro-preview-customtools:generateContent` — Google Gemini (`x-goog-api-key: $AIHUBMIX_API_KEY`; streaming: `:streamGenerateContent?alt=sse`) - `POST /v1/messages` — Anthropic Messages (`x-api-key: $AIHUBMIX_API_KEY` + `anthropic-version: 2023-06-01`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-3.1-pro-preview-customtools","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gemini-3.1-pro-preview-customtools","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gemini-3.1-pro-preview-customtools - Try in browser: https://aihubmix.com/playground?model=gemini-3.1-pro-preview-customtools - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gemini-3.1-pro-preview-customtools.983b2677.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-3.1-pro-preview-customtools/llms.txt --- # Gemini 3.1 Pro Preview Search Model id on AIHubMix: `gemini-3.1-pro-preview-search` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Gemini-3.1-pro-preview-search integrates Google's official search functionality; the search feature incurs an additional separate fee log directly incorporated into the scoring, but the log details are not displayed; this will be fixed in the future to show the details; it only supports OpenAI-compatible format calls and does not support the Gemini SDK; for the Gemini native SDK, please directly set the official search parameters. - Developer: Google - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio, PDF - Capabilities: Thinking, Tool calling, Web search, URL context, Code interpreter, File search, Structured outputs, Prompt caching - Release date: 2026-02-19 - Pricing: $2/M input tokens, $12/M output tokens, $0.5/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-3.1-pro-preview-search","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gemini-3.1-pro-preview-search","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gemini-3.1-pro-preview-search - Try in browser: https://aihubmix.com/playground?model=gemini-3.1-pro-preview-search - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gemini-3.1-pro-preview-search.994a2c70.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-3.1-pro-preview-search/llms.txt --- # GPT 5.4 Mini Model id on AIHubMix: `gpt-5.4-mini` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-5.4 mini is a faster, more efficient model that inherits the advantages of GPT-5.4 and is specifically optimized for high-volume workloads. - Developer: OpenAI - Context window: 400,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Computer use, File search, Structured outputs, Prompt caching - Release date: 2026-03-17 - Pricing: $0.75/M input tokens, $4.5/M output tokens, $0.075/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/responses` — OpenAI Responses (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/messages` — Anthropic Messages (`x-api-key: $AIHUBMIX_API_KEY` + `anthropic-version: 2023-06-01`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-5.4-mini","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-5.4-mini","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-5.4-mini - Try in browser: https://aihubmix.com/playground?model=gpt-5.4-mini - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-5.4-mini.9057df1e.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-5.4-mini/llms.txt --- # GPT 5.4 Nano Model id on AIHubMix: `gpt-5.4-nano` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-5.4 nano is designed for tasks where speed and cost are most important, such as classification, data extraction, ranking, and sub-agent scenarios. - Developer: OpenAI - Context window: 400,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, File search, Structured outputs, Prompt caching - Release date: 2026-03-17 - Pricing: $0.2/M input tokens, $1.25/M output tokens, $0.02/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-5.4-nano","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-5.4-nano","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-5.4-nano - Try in browser: https://aihubmix.com/playground?model=gpt-5.4-nano - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-5.4-nano.e341ed79.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-5.4-nano/llms.txt --- # Qwen3.5 Plus Model id on AIHubMix: `qwen3.5-plus` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > The Qwen 3.5 native vision-language Plus model is built on a hybrid architecture that integrates linear attention mechanisms with sparse mixture-of-experts models, achieving higher inference efficiency. In multiple task evaluations, the 3.5 series has demonstrated outstanding performance comparable to current leading frontier models, with leapfrog improvements over the 3 series in both pure-text and multimodal capabilities. This model version is functionally equivalent to the snapshot model qwen3.5-plus-2026-02-15. - Developer: Qwen - Context window: 1,000,000 tokens - Input modalities: text, image, video - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Structured outputs, Prompt caching - Release date: 2026-02-16 - Pricing: $0.1096/M input tokens, $0.6576/M output tokens, $0.011/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen3.5-plus","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"qwen3.5-plus","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/qwen3.5-plus - Try in browser: https://aihubmix.com/playground?model=qwen3.5-plus - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/qwen3.5-plus.ae13d5fb.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen3.5-plus/llms.txt --- # Claude Sonnet 4.6 Thinking Model id on AIHubMix: `claude-sonnet-4-6-think` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Claude sonnet 4.6 does not enable reasoning mode by default. To access its deep reasoning capabilities, users would typically need to call the native Claude API. To make this capability available through an OpenAI-compatible interface, we provide the claude-sonnet-4-6-think model, which has reasoning mode pre-enabled and a default 32k-token context window, allowing it to be called directly via the OpenAI unified API. Claude sonnet 4.5 Think is a reasoning-focused variant of Claude sonnet 4.6 designed for advanced tasks that require rigorous reasoning, complex decision-making, and long-chain analysis. Aside from its enhanced reasoning mechanism, all other capabilities remain consistent with the standard Claude sonnet 4.6 model, making it well suited for complex engineering problem decomposition, multi-stage planning, and logic-intensive analysis. - Developer: Anthropic - Context window: 1,000,000 tokens - Input modalities: text, image, PDF - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Computer use, Memory tool, Structured outputs, Citations, Prompt caching - Release date: 2026-02-17 - Pricing: $3/M input tokens, $15/M output tokens, $0.3/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"claude-sonnet-4-6-think","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"claude-sonnet-4-6-think","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/claude-sonnet-4-6-think - Try in browser: https://aihubmix.com/playground?model=claude-sonnet-4-6-think - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/claude-sonnet-4-6-think.01a8c5cc.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/claude-sonnet-4-6-think/llms.txt --- # GPT 5.3 Chat Model id on AIHubMix: `gpt-5.3-chat-latest` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-5.3Chat refers to the GPT-5.3 snapshot currently used in ChatGPT and is optimized for conversational use cases. While GPT-5.2 is recommended for most API applications, GPT-5.3Chat is ideal for testing the latest improvements in chat-based interactions. - Developer: OpenAI - Context window: 128,000 tokens - Input modalities: text, image - Capabilities: Streaming, Tool calling, Web search, Code interpreter, File search, Structured outputs, Prompt caching - Release date: 2026-03-03 - Pricing: $1.75/M input tokens, $14/M output tokens, $0.175/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-5.3-chat-latest","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-5.3-chat-latest","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-5.3-chat-latest - Try in browser: https://aihubmix.com/playground?model=gpt-5.3-chat-latest - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-5.3-chat-latest.881b4d7c.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-5.3-chat-latest/llms.txt --- # GPT-5.3-Codex Model id on AIHubMix: `gpt-5.3-codex` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-5.3-Codex is optimized for agentic coding tasks in Codex or similar environments. GPT-5.3-Codex supports low, medium, high, and xhigh reasoning effort settings. If you want to learn more about prompting GPT-5.3-Codex, refer to our dedicated guide. - Developer: OpenAI - Context window: 400,000 tokens - Input modalities: text, image, PDF - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Release date: 2026-02-05 - Pricing: $1.75/M input tokens, $14/M output tokens, $0.175/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-5.3-codex","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-5.3-codex","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-5.3-codex - Try in browser: https://aihubmix.com/playground?model=gpt-5.3-codex - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-5.3-codex.48631aac.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-5.3-codex/llms.txt --- # Qwen3.5 122B A10B Model id on AIHubMix: `qwen3.5-122b-a10b` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > The Qwen 3.5 native vision-language Plus model is built on a hybrid architecture that integrates linear attention mechanisms with sparse mixture-of-experts models, achieving higher inference efficiency. In multiple task evaluations, the 3.5 series has demonstrated outstanding performance comparable to current leading frontier models, with leapfrog improvements over the 3 series in both pure-text and multimodal capabilities. This model version is functionally equivalent to the snapshot model qwen3.5-plus-2026-02-15. - Developer: Qwen - Context window: 262,144 tokens - Input modalities: text, image, video - Capabilities: Thinking, Tool calling, Web search, Code interpreter, Structured outputs - Release date: 2026-02-23 - Pricing: $0.1126/M input tokens, $0.9008/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen3.5-122b-a10b","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"qwen3.5-122b-a10b","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/qwen3.5-122b-a10b - Try in browser: https://aihubmix.com/playground?model=qwen3.5-122b-a10b - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/qwen3.5-122b-a10b.366291e7.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen3.5-122b-a10b/llms.txt --- # Qwen3.5 27B Model id on AIHubMix: `qwen3.5-27b` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > The Qwen 3.5 native vision-language Plus model is built on a hybrid architecture that integrates linear attention mechanisms with sparse mixture-of-experts models, achieving higher inference efficiency. In multiple task evaluations, the 3.5 series has demonstrated outstanding performance comparable to current leading frontier models, with leapfrog improvements over the 3 series in both pure-text and multimodal capabilities. This model version is functionally equivalent to the snapshot model qwen3.5-plus-2026-02-15. - Developer: Qwen - Context window: 262,144 tokens - Input modalities: text, image, video - Capabilities: Thinking, Tool calling, Web search, Code interpreter, Structured outputs - Release date: 2026-02-23 - Pricing: $0.0846/M input tokens, $0.6768/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen3.5-27b","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"qwen3.5-27b","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/qwen3.5-27b - Try in browser: https://aihubmix.com/playground?model=qwen3.5-27b - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/qwen3.5-27b.98f6b2fd.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen3.5-27b/llms.txt --- # Qwen3.5 35B A3B Model id on AIHubMix: `qwen3.5-35b-a3b` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > The Qwen 3.5 native vision-language Plus model is built on a hybrid architecture that integrates linear attention mechanisms with sparse mixture-of-experts models, achieving higher inference efficiency. In multiple task evaluations, the 3.5 series has demonstrated outstanding performance comparable to current leading frontier models, with leapfrog improvements over the 3 series in both pure-text and multimodal capabilities. This model version is functionally equivalent to the snapshot model qwen3.5-plus-2026-02-15. - Developer: Qwen - Context window: 262,144 tokens - Input modalities: text, image, video - Capabilities: Thinking, Tool calling, Web search, Code interpreter, Structured outputs - Release date: 2026-02-23 - Pricing: $0.0564/M input tokens, $0.4512/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen3.5-35b-a3b","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"qwen3.5-35b-a3b","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/qwen3.5-35b-a3b - Try in browser: https://aihubmix.com/playground?model=qwen3.5-35b-a3b - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/qwen3.5-35b-a3b.1e22a770.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen3.5-35b-a3b/llms.txt --- # Qwen3.5 397B A17B Model id on AIHubMix: `qwen3.5-397b-a17b` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > The Qwen 3.5 native vision-language Plus model is built on a hybrid architecture that integrates linear attention mechanisms with sparse mixture-of-experts models, achieving higher inference efficiency. In multiple task evaluations, the 3.5 series has demonstrated outstanding performance comparable to current leading frontier models, with leapfrog improvements over the 3 series in both pure-text and multimodal capabilities. This model version is functionally equivalent to the snapshot model qwen3.5-plus-2026-02-15. - Developer: Qwen - Context window: 262,144 tokens - Input modalities: text, image, video - Capabilities: Thinking, Tool calling, Web search, Code interpreter, Structured outputs - Release date: 2026-02-15 - Pricing: $0.1644/M input tokens, $0.9864/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen3.5-397b-a17b","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"qwen3.5-397b-a17b","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/qwen3.5-397b-a17b - Try in browser: https://aihubmix.com/playground?model=qwen3.5-397b-a17b - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/qwen3.5-397b-a17b.97b53fd4.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen3.5-397b-a17b/llms.txt --- # Qwen3.5 Flash Model id on AIHubMix: `qwen3.5-flash` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > The Qwen3.5 native vision-language Flash series models are designed with a hybrid architecture that integrates linear attention mechanisms and sparse mixture-of-experts models, achieving higher inference efficiency. Compared with the 3 series, the models deliver leapfrog improvements in both pure-text and multimodal performance; they respond quickly and combine inference speed with high performance. - Developer: Qwen - Context window: 1,000,000 tokens - Input modalities: text, image, video - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Structured outputs, Prompt caching - Release date: 2026-02-23 - Pricing: $0.0282/M input tokens, $0.282/M output tokens, $0.0028/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen3.5-flash","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"qwen3.5-flash","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/qwen3.5-flash - Try in browser: https://aihubmix.com/playground?model=qwen3.5-flash - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/qwen3.5-flash.68800fa1.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen3.5-flash/llms.txt --- # Coding GLM 5.1 Model id on AIHubMix: `coding-glm-5.1` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Only supports OpenAI-compatible formats. - Developer: Z.AI - Context window: 200,000 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2026-04-11 - Pricing: $0.06/M input tokens, $0.22/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"coding-glm-5.1","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"coding-glm-5.1","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/coding-glm-5.1 - Try in browser: https://aihubmix.com/playground?model=coding-glm-5.1 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/coding-glm-5.1.afdfab65.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/coding-glm-5.1/llms.txt --- # Doubao Seed 2.0 Pro Model id on AIHubMix: `doubao-seed-2-0-pro` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Doubao flagship all-purpose general model, targeting complex reasoning and long-chain task execution scenarios in the Agent era. It emphasizes multimodal understanding, long-context reasoning, structured generation, and tool-augmented execution. It excels at handling complex instructions and multi-constraint execution, reliably addressing multi-step complex planning, intricate image-text reasoning, video content understanding, and high-difficulty analysis scenarios. - Developer: ByteDance - Context window: 256,000 tokens - Input modalities: text, image, video, document - Capabilities: Thinking, Tool calling, Web search, File search, Document input, Prompt caching, Server-side sessions - Release date: 2026-02-14 - Pricing: $0.4958/M input tokens, $2.479/M output tokens, $0.0992/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"doubao-seed-2-0-pro","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"doubao-seed-2-0-pro","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/doubao-seed-2-0-pro - Try in browser: https://aihubmix.com/playground?model=doubao-seed-2-0-pro - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/doubao-seed-2-0-pro.5dcd4ee0.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/doubao-seed-2-0-pro/llms.txt --- # GPT 5.4 High Model id on AIHubMix: `gpt-5.4-high` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-5.4 supports configurable reasoning effort only through the /responses endpoint. To make higher-intensity reasoning available directly via the /chat interface, GPT-5.4-High is provided as a reasoning-enhanced variant of GPT-5.4 with reasoning_effort preset to high. It is designed for tasks that require deeper analysis, stronger result consistency, and greater controllability. By applying more aggressive reasoning strategies and more effective use of extended context, the model delivers clearer and more reliable responses, making it well suited for complex agent workflows, long-chain decision-making, and reliability-critical advanced applications. - Developer: OpenAI - Context window: 1,050,000 tokens - Input modalities: text, image, PDF - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Computer use, File search, Structured outputs, Prompt caching, Server-side sessions - Pricing: $2.5/M input tokens, $15/M output tokens, $0.25/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-5.4-high","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-5.4-high","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-5.4-high - Try in browser: https://aihubmix.com/playground?model=gpt-5.4-high - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-5.4-high.d99bd0c3.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-5.4-high/llms.txt --- # GPT 5.4 Low Model id on AIHubMix: `gpt-5.4-low` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-5.4 supports configuring reasoning strength only through the /responses endpoint. To make lower-overhead reasoning available directly in the /chat endpoint, the GPT-5.4-Low model is provided. This model is based on GPT-5.4 with reasoning_effort preset to low. This model is designed for use cases that are sensitive to response latency and cost. By adopting a lighter reasoning strategy, it delivers stable responses with lower latency and higher throughput. It is well suited for high-concurrency conversations, real-time interactions, basic Q&A, and scenarios where deep reasoning is not required. - Developer: OpenAI - Context window: 1,050,000 tokens - Input modalities: text, image, PDF - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Computer use, File search, Structured outputs, Prompt caching, Server-side sessions - Pricing: $2.5/M input tokens, $15/M output tokens, $0.25/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-5.4-low","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-5.4-low","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-5.4-low - Try in browser: https://aihubmix.com/playground?model=gpt-5.4-low - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-5.4-low.48f92d49.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-5.4-low/llms.txt --- # GPT 5.4 Pro Model id on AIHubMix: `gpt-5.4-pro` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Please note: this model is extremely expensive and very slow. If a request fails due to network issues, you may still be charged heavily; we cannot refund charges incurred by requests to this model. GPT-5.4 pro is available in the Responses API only to enable support for multi-turn model interactions before responding to API requests, and other advanced API features in the future. Since GPT-5.4 pro is designed to tackle tough problems, some requests may take several minutes to finish. To avoid timeout, please set a longer timeout duration. It is recommended to use this under good network conditions. - Developer: OpenAI - Context window: 1,050,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Web search, Computer use, File search, Background mode - Release date: 2026-03-05 - Pricing: $30/M input tokens, $180/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-5.4-pro","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-5.4-pro","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-5.4-pro - Try in browser: https://aihubmix.com/playground?model=gpt-5.4-pro - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-5.4-pro.b3480c6c.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-5.4-pro/llms.txt --- # Qwen3 Coder Next Model id on AIHubMix: `qwen3-coder-next` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > The Qwen3 series is a next-generation code-generation model with results close to Qwen3-Coder-Plus while offering superior performance. The model is optimized for repository-level understanding, supports multi-turn tool interactions, and improves compatibility with agentic coding tools. - Developer: Qwen - Context window: 262,144 tokens - Input modalities: text - Release date: 2026-02-03 - Pricing: $0.137/M input tokens, $0.548/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen3-coder-next","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"qwen3-coder-next","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/qwen3-coder-next - Try in browser: https://aihubmix.com/playground?model=qwen3-coder-next - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/qwen3-coder-next.6dccc24c.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen3-coder-next/llms.txt --- # Claude Opus 4.6 Model id on AIHubMix: `claude-opus-4-6` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Claude Opus 4.6 is Anthropic’s latest state-of-the-art reasoning model. It features an adaptive “thinking” mode that dynamically decides when to think and how much to think. At the default effort level (high), Claude will almost always engage in thinking. At lower effort levels, it may skip thinking for simple problems. ⚠️ The minimum cache token for claude-opus-4-6 has been increased from 1,024 to 4,096 tokens. - Developer: Anthropic - Context window: 1,000,000 tokens - Input modalities: text, image, PDF - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Computer use, Memory tool, Structured outputs, Citations, Prompt caching - Release date: 2026-02-05 - Pricing: $5/M input tokens, $25/M output tokens, $0.5/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"claude-opus-4-6","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"claude-opus-4-6","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/claude-opus-4-6 - Try in browser: https://aihubmix.com/playground?model=claude-opus-4-6 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/claude-opus-4-6.04927ea3.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/claude-opus-4-6/llms.txt --- # Coding GLM 5.1 (free) Model id on AIHubMix: `coding-glm-5.1-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > coding-glm-5.1-free is the open and free version of coding-glm-5.1. To maintain reliable service, each account is limited to 5 requests per minute, 100 requests per day, and 1 million tokens per day. - Developer: Z.AI - Context window: 200,000 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2026-04-11 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"coding-glm-5.1-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"coding-glm-5.1-free","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/coding-glm-5.1-free - Try in browser: https://aihubmix.com/playground?model=coding-glm-5.1-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/coding-glm-5.1-free.de200b86.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/coding-glm-5.1-free/llms.txt --- # Coding MiniMax M2.7 (free) Model id on AIHubMix: `coding-minimax-m2.7-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > coding-minimax-m2.7-free is a free and open version offered by AIHubMix specifically for MiniMax users. To maintain reliable service, each account is limited to 5 requests per minute, 100 requests per day, and 1 million tokens per day. - Developer: Minimax - Context window: 204,800 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Prompt caching - Release date: 2026-03-18 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"coding-minimax-m2.7-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"coding-minimax-m2.7-free","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/coding-minimax-m2.7-free - Try in browser: https://aihubmix.com/playground?model=coding-minimax-m2.7-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/coding-minimax-m2.7-free.6d286bf9.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/coding-minimax-m2.7-free/llms.txt --- # GLM 5 Vision Turbo Model id on AIHubMix: `glm-5v-turbo` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GLM-5V-Turbo is Zhipu's first multimodal coding foundation model, built for visual programming tasks. It can natively handle multimodal inputs such as images, videos, and text, and is adept at long-horizon planning, complex programming, and action execution; deeply adapted to Agent workflows, it can collaborate closely with agents like Claude Code and OpenClaw to complete the full closed loop of "understand the environment → plan actions → execute tasks." - Developer: Z.AI - Context window: 200,000 tokens - Input modalities: text, image, video, document - Capabilities: Thinking, Streaming, Document input, Prompt caching - Release date: 2026-04-01 - Pricing: $0.7042/M input tokens, $3.0985/M output tokens, $0.169/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"glm-5v-turbo","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"glm-5v-turbo","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/glm-5v-turbo - Try in browser: https://aihubmix.com/playground?model=glm-5v-turbo - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/glm-5v-turbo.a68d5474.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/glm-5v-turbo/llms.txt --- # MiniMax M2.7 Model id on AIHubMix: `minimax-m2.7` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > MiniMax M2.7 can autonomously build complex Agent Harnesses and, leveraging capabilities such as Agent Teams, complex Skills, and the Tool Search tool, complete highly complex productivity tasks. - Developer: Minimax - Context window: 204,800 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Prompt caching - Release date: 2026-03-18 - Pricing: $0.2958/M input tokens, $1.1832/M output tokens, $0.0592/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"minimax-m2.7","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"minimax-m2.7","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/minimax-m2.7 - Try in browser: https://aihubmix.com/playground?model=minimax-m2.7 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/minimax-m2.7.1361912b.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/minimax-m2.7/llms.txt --- # Claude Opus 4.6 Thinking Model id on AIHubMix: `claude-opus-4-6-think` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Claude Opus 4.6 does not enable reasoning mode by default. To access its deep reasoning capabilities, users would typically need to call the native Claude API. To make this capability available through an OpenAI-compatible interface, we provide the claude-opus-4-6-think model, which has reasoning mode pre-enabled and a default 32k-token context window, allowing it to be called directly via the OpenAI unified API. Claude Opus 4.5 Think is a reasoning-focused variant of Claude Opus 4.6 designed for advanced tasks that require rigorous reasoning, complex decision-making, and long-chain analysis. Aside from its enhanced reasoning mechanism, all other capabilities remain consistent with the standard Claude Opus 4.6 model, making it well suited for complex engineering problem decomposition, multi-stage planning, and logic-intensive analysis. - Developer: Anthropic - Context window: 1,000,000 tokens - Input modalities: text, image, PDF - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Computer use, Memory tool, Structured outputs, Citations, Prompt caching - Release date: 2026-02-05 - Pricing: $5/M input tokens, $25/M output tokens, $0.5/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"claude-opus-4-6-think","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"claude-opus-4-6-think","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/claude-opus-4-6-think - Try in browser: https://aihubmix.com/playground?model=claude-opus-4-6-think - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/claude-opus-4-6-think.8aba16e1.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/claude-opus-4-6-think/llms.txt --- # Coding GLM 5 (free) Model id on AIHubMix: `coding-glm-5-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > coding-glm-5-free is the open and free version of coding-glm-5. To ensure stable service performance, usage limits are in place: up to 5 requests per minute, 100 requests per day, and a daily token allowance of 1 million. - Developer: Z.AI - Context window: 200,000 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2026-02-11 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"coding-glm-5-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"coding-glm-5-free","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/coding-glm-5-free - Try in browser: https://aihubmix.com/playground?model=coding-glm-5-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/coding-glm-5-free.d6c2dd7d.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/coding-glm-5-free/llms.txt --- # Coding GLM 5 Turbo (free) Model id on AIHubMix: `coding-glm-5-turbo-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > coding-glm-5-turbo-free is the open and free version of coding-glm-5-turbo. To ensure stable service performance, usage limits are in place: up to 5 requests per minute, 100 requests per day, and a daily token allowance of 1 million. - Developer: Z.AI - Context window: 204,800 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2026-03-16 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"coding-glm-5-turbo-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"coding-glm-5-turbo-free","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/coding-glm-5-turbo-free - Try in browser: https://aihubmix.com/playground?model=coding-glm-5-turbo-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/coding-glm-5-turbo-free.f8344c85.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/coding-glm-5-turbo-free/llms.txt --- # Coding MiniMax M2.5 (free) Model id on AIHubMix: `coding-minimax-m2.5-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > coding-minimax-m2.5-free is a free and open version offered by AIHubMix specifically for MiniMax users. To maintain stable service operations, the following usage limits apply: a maximum of 5 requests per minute, 100 total requests per day, and a daily quota of 1 million tokens. - Developer: Minimax - Context window: 204,800 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Prompt caching - Release date: 2026-02-12 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"coding-minimax-m2.5-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"coding-minimax-m2.5-free","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/coding-minimax-m2.5-free - Try in browser: https://aihubmix.com/playground?model=coding-minimax-m2.5-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/coding-minimax-m2.5-free.868a5b2b.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/coding-minimax-m2.5-free/llms.txt --- # Doubao Seed 2.0 Code Preview Model id on AIHubMix: `doubao-seed-2-0-code-preview` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > The Doubao 2.0 series is a coding model optimized for real programming environments, capable of reliably invoking tools in common IDEs such as Claude Code. The model is specially optimized for frontend capabilities and performs well with common frontend frameworks. The model supports using Skills and can work with various custom skills. - Developer: Doubao - Context window: 256,000 tokens - Input modalities: text, image, video, document - Capabilities: Thinking, Tool calling, Web search, File search, Document input, Prompt caching, Server-side sessions - Release date: 2026-02-14 - Pricing: $0.4822/M input tokens, $2.411/M output tokens, $0.0964/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"doubao-seed-2-0-code-preview","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"doubao-seed-2-0-code-preview","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/doubao-seed-2-0-code-preview - Try in browser: https://aihubmix.com/playground?model=doubao-seed-2-0-code-preview - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/doubao-seed-2-0-code-preview.67b52e39.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/doubao-seed-2-0-code-preview/llms.txt --- # Doubao Seed 2.0 Lite 260215 Model id on AIHubMix: `doubao-seed-2-0-lite-260215` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Doubao Coding model optimized for real-world programming environments that can reliably invoke tools in common IDEs such as Claude Code. The model is specially optimized for frontend capabilities and performs well with common frontend frameworks. The model supports Skills and can work with various custom skills. - Developer: Doubao - Context window: 256,000 tokens - Input modalities: text, image, video, document - Capabilities: Thinking, Tool calling, Web search, File search, Document input, Structured outputs, Prompt caching, Server-side sessions - Release date: 2026-02-15 - Pricing: $0.0904/M input tokens, $0.5425/M output tokens, $0.0181/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"doubao-seed-2-0-lite-260215","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"doubao-seed-2-0-lite-260215","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/doubao-seed-2-0-lite-260215 - Try in browser: https://aihubmix.com/playground?model=doubao-seed-2-0-lite-260215 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/doubao-seed-2-0-lite-260215.506cb1d5.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/doubao-seed-2-0-lite-260215/llms.txt --- # Gemini 3 Flash Preview Model id on AIHubMix: `gemini-3-flash-preview` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > gemini-3-flash-preview is Google's latest released, most balanced model, excelling in speed, scale, and cutting-edge intelligence. - Developer: Google - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio, PDF - Capabilities: Thinking, Tool calling, Web search, URL context, Code interpreter, Computer use, File search, Structured outputs, Prompt caching - Release date: 2025-12-17 - Pricing: $0.5/M input tokens, $3/M output tokens, $0.05/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-3-flash-preview","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gemini-3-flash-preview","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gemini-3-flash-preview - Try in browser: https://aihubmix.com/playground?model=gemini-3-flash-preview - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gemini-3-flash-preview.f8311517.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-3-flash-preview/llms.txt --- # Gemini 3 Flash Preview Search Model id on AIHubMix: `gemini-3-flash-preview-search` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Gemini-3-flash-preview-search integrates Google's official search functionality; the search feature incurs an additional separate fee log directly incorporated into the scoring, but the log details are not displayed; this will be fixed in the future to show the details; it only supports OpenAI-compatible format calls and does not support the Gemini SDK; for the Gemini native SDK, please directly set the official search parameters. - Developer: Google - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio, PDF - Capabilities: Thinking, Tool calling, Web search, URL context, Code interpreter, Computer use, File search, Structured outputs, Prompt caching - Release date: 2025-12-17 - Pricing: $0.5/M input tokens, $3/M output tokens, $0.05/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-3-flash-preview-search","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gemini-3-flash-preview-search","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gemini-3-flash-preview-search - Try in browser: https://aihubmix.com/playground?model=gemini-3-flash-preview-search - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gemini-3-flash-preview-search.611a90c9.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-3-flash-preview-search/llms.txt --- # GLM 5 Turbo Model id on AIHubMix: `glm-5-turbo` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GLM-5-Turbo is a foundational model deeply optimized for the OpenClaw scenario. From the training stage it has been specifically optimized for the core requirements of OpenClaw tasks, enhancing key capabilities such as tool invocation, instruction following, scheduled and persistent tasks, and long-chain execution. - Developer: Z.AI - Context window: 204,800 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2026-03-16 - Pricing: $1.2/M input tokens, $3.9996/M output tokens, $0.24/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"glm-5-turbo","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"glm-5-turbo","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/glm-5-turbo - Try in browser: https://aihubmix.com/playground?model=glm-5-turbo - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/glm-5-turbo.5e0c6315.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/glm-5-turbo/llms.txt --- # CC GLM 5.1 Model id on AIHubMix: `cc-glm-5.1` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Supports Claude native interface, can be directly requested in Claude Code. - Developer: Z.AI - Context window: 200,000 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Pricing: $0.06/M input tokens, $0.22/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"cc-glm-5.1","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"cc-glm-5.1","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/cc-glm-5.1 - Try in browser: https://aihubmix.com/playground?model=cc-glm-5.1 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/cc-glm-5.1.a9191638.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/cc-glm-5.1/llms.txt --- # Claude Opus 4.5 Model id on AIHubMix: `claude-opus-4-5` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Claude Opus 4.5 is Anthropic’s latest frontier reasoning model, optimized for complex engineering, agentic workflows, and long-horizon computer use. It features strong multimodal capabilities, improved resistance to prompt injection, and a new Verbosity parameter to control token efficiency. With advanced tool use, extended context, and multi-agent support, Opus 4.5 excels in autonomous research, debugging, planning, and spreadsheet/browser operations. ⚠️ The minimum cache token for claude-opus-4-5 has been increased from 1,024 to 4,096 tokens. - Developer: Anthropic - Context window: 200,000 tokens - Input modalities: text, image, PDF - Capabilities: Thinking, Streaming, Tool calling, Code interpreter, Computer use, Memory tool, Structured outputs, Citations, Prompt caching - Release date: 2025-11-24 - Pricing: $5/M input tokens, $25/M output tokens, $0.5/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"claude-opus-4-5","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"claude-opus-4-5","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/claude-opus-4-5 - Try in browser: https://aihubmix.com/playground?model=claude-opus-4-5 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/claude-opus-4-5.2680b112.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/claude-opus-4-5/llms.txt --- # Claude Opus 4.5 Thinking Model id on AIHubMix: `claude-opus-4-5-think` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Claude Opus 4.5 does not enable reasoning mode by default. To access its deep reasoning capabilities, users would typically need to call the native Claude API. To make this capability available through an OpenAI-compatible interface, we provide the claude-opus-4-5-think model, which has reasoning mode pre-enabled and a default 32k-token context window, allowing it to be called directly via the OpenAI unified API. Claude Opus 4.5 Think is a reasoning-focused variant of Claude Opus 4.5 designed for advanced tasks that require rigorous reasoning, complex decision-making, and long-chain analysis. Aside from its enhanced reasoning mechanism, all other capabilities remain consistent with the standard Claude Opus 4.5 model, making it well suited for complex engineering problem decomposition, multi-stage planning, and logic-intensive analysis. - Developer: Anthropic - Context window: 200,000 tokens - Input modalities: text, image, PDF - Capabilities: Thinking, Streaming, Tool calling, Code interpreter, Computer use, Memory tool, Structured outputs, Citations, Prompt caching - Release date: 2025-11-24 - Pricing: $5/M input tokens, $25/M output tokens, $0.5/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"claude-opus-4-5-think","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"claude-opus-4-5-think","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/claude-opus-4-5-think - Try in browser: https://aihubmix.com/playground?model=claude-opus-4-5-think - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/claude-opus-4-5-think.e9733d65.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/claude-opus-4-5-think/llms.txt --- # ERNIE Image Turbo Model id on AIHubMix: `ernie-image-turbo` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > The Ernie-image-Turbo model is an 8-step distilled version of the Ernie-image model, also with 8 billion parameters, offering a 6x speedup compared to pre-distillation, and is suitable for low-latency, local/on-device scenarios. - Developer: Baidu - Pricing: per-generation — see https://aihubmix.com/model/ernie-image-turbo ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"ernie-image-turbo","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `model` | string | yes | | | | `output_format` | string,null | no | "png" | png, jpeg, webp, null | | `prompt` | string | yes | | | | `response_format` | string,null | no | "url" | url, b64_json, null | | `size` | string,null | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/ernie-image-turbo/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/ernie-image-turbo/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/ernie-image-turbo - Try in browser: https://aihubmix.com/playground?model=ernie-image-turbo - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/ernie-image-turbo/llms.txt --- # Xing4.0 29B A4B (free) Model id on AIHubMix: `xing4.0-29b-a4b-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Xing4.0-29B-A4B is an open-weight mixture-of-experts (MoE) language model developed by Xingchen Agi and China Telecom's AI team (TeleAI) as the successor to the TeleChat series. It features a total of 29 billion parameters, with 4 billion active per token, to deliver highly efficient performance. With an expansive context length of 262,144 tokens, this model is well-suited for processing long-form documents and complex tasks. - Developer: Xingchen Agi - Context window: 262,144 tokens - Input modalities: text - Capabilities: Thinking, Tool calling - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"xing4.0-29b-a4b-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"xing4.0-29b-a4b-free","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/xing4.0-29b-a4b-free - Try in browser: https://aihubmix.com/playground?model=xing4.0-29b-a4b-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/xing4.0-29b-a4b-free.685d67a6.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/xing4.0-29b-a4b-free/llms.txt --- # Cohere Command A Model id on AIHubMix: `cohere-command-a` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Command A is Cohere most performant model to date, excelling at tool use, agents, retrieval augmented generation (RAG), and multilingual use cases. Command A has a context length of 256K, only requires two GPUs to run, and has 150% higher throughput compared to Command R+ 08-2024. - Developer: Cohere - Context window: 256,000 tokens - Input modalities: text - Capabilities: Tool calling, Structured outputs - Release date: 2025-03-13 - Pricing: $2.5/M input tokens, $10/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"cohere-command-a","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"cohere-command-a","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/cohere-command-a - Try in browser: https://aihubmix.com/playground?model=cohere-command-a - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/cohere-command-a.447a58d0.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/cohere-command-a/llms.txt --- # CC MiniMax M3 Model id on AIHubMix: `cc-minimax-m3` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > For Claude Code only - Developer: Minimax - Context window: 1,000,000 tokens - Input modalities: text, image, video - Capabilities: Thinking, Streaming, Tool calling, Web search, Prompt caching - Pricing: $0.1/M input tokens, $0.1/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"cc-minimax-m3","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"cc-minimax-m3","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/cc-minimax-m3 - Try in browser: https://aihubmix.com/playground?model=cc-minimax-m3 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/cc-minimax-m3.a3728862.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/cc-minimax-m3/llms.txt --- # Coding MiniMax M3 Model id on AIHubMix: `coding-minimax-m3` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms - Developer: Minimax - Context window: 1,000,000 tokens - Input modalities: text, image, video - Capabilities: Thinking, Streaming, Tool calling, Web search, Prompt caching - Release date: 2026-06-01 - Pricing: $0.2/M input tokens, $0.2/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"coding-minimax-m3","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"coding-minimax-m3","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/coding-minimax-m3 - Try in browser: https://aihubmix.com/playground?model=coding-minimax-m3 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/coding-minimax-m3.8bddda96.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/coding-minimax-m3/llms.txt --- # Coding GLM 5 Model id on AIHubMix: `coding-glm-5` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Only supports OpenAI-compatible formats. - Developer: Z.AI - Context window: 200,000 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2026-02-12 - Pricing: $0.06/M input tokens, $0.22/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"coding-glm-5","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"coding-glm-5","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/coding-glm-5 - Try in browser: https://aihubmix.com/playground?model=coding-glm-5 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/coding-glm-5.c24b054a.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/coding-glm-5/llms.txt --- # Coding GLM 5 Turbo Model id on AIHubMix: `coding-glm-5-turbo` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Only supports OpenAI-compatible formats. - Developer: Z.AI - Context window: 204,800 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2026-03-16 - Pricing: $0.06/M input tokens, $0.22/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"coding-glm-5-turbo","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"coding-glm-5-turbo","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/coding-glm-5-turbo - Try in browser: https://aihubmix.com/playground?model=coding-glm-5-turbo - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/coding-glm-5-turbo.8c65444a.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/coding-glm-5-turbo/llms.txt --- # GLM 4.7 Model id on AIHubMix: `glm-4.7` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GLM-4.7 is Zhiyuan's latest flagship model. GLM-4.7 enhances coding capabilities, long-range task planning, and tool collaboration for Agentic Coding scenarios, achieving leading performance among open-source models on several current public benchmarks. It features improved general capabilities, with responses that are more concise and natural, and writing that is more immersive. When executing complex agent tasks and tool usage, it follows instructions more strictly, with further improvements in the frontend aesthetics of Artifacts and Agentic Coding as well as the efficiency of completing long-range tasks. - Developer: Z.AI - Context window: 200,000 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2025-12-22 - Pricing: $0.274/M input tokens, $1.0959/M output tokens, $0.0548/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"glm-4.7","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"glm-4.7","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/glm-4.7 - Try in browser: https://aihubmix.com/playground?model=glm-4.7 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/glm-4.7.c2aad54a.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/glm-4.7/llms.txt --- # GLM 4.7 Flash (free) Model id on AIHubMix: `glm-4.7-flash-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > The glm-4.7-flash free model has usage restrictions to ensure stable service operation: a maximum of 5 requests per minute, no more than 500 requests per day, and a daily usage quota of 1 million tokens. - Developer: Z.AI - Context window: 200,000 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2026-01-19 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"glm-4.7-flash-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"glm-4.7-flash-free","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/glm-4.7-flash-free - Try in browser: https://aihubmix.com/playground?model=glm-4.7-flash-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/glm-4.7-flash-free.43f6147f.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/glm-4.7-flash-free/llms.txt --- # Coding GLM 4.7 (free) Model id on AIHubMix: `coding-glm-4.7-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > coding-glm-4.7-free is the open and free version of coding-glm-4.7. To ensure stable service performance, usage limits are in place: up to 5 requests per minute, 500 requests per day, and a daily token allowance of 1 million. - Developer: Z.AI - Context window: 200,000 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2025-12-22 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"coding-glm-4.7-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"coding-glm-4.7-free","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/coding-glm-4.7-free - Try in browser: https://aihubmix.com/playground?model=coding-glm-4.7-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/coding-glm-4.7-free.86e8d879.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/coding-glm-4.7-free/llms.txt --- # Gemini 3 Pro Image Preview Model id on AIHubMix: `gemini-3-pro-image-preview` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model is scheduled for retirement.** Calls still work until then. Plan to migrate to `gemini-3-pro-image`. Its guide: https://aihubmix.com/model/gemini-3-pro-image/llms.txt > > Once retired, requests return HTTP 404 `model_retired`. Responses already carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > Gemini-3-Pro-Image-Preview (Nano Banana Pro) is a high-performance image generation and editing model built on Gemini 3 Pro. It delivers enhanced multimodal understanding and real-world semantic reasoning, enabling fast creation of well-structured visual content such as infographics, product sketches, and multi-subject scenes. It can also leverage real-time knowledge through Search grounding. The model excels in text rendering, consistent multi-image blending, and identity preservation, while offering fine-grained creative controls like localized edits, lighting and focus adjustments, camera transformations, and flexible aspect ratios. It’s ideal for rapid design, concept previews, product visualization, and everyday image generation workflows. - Developer: Google - Input modalities: text, image - Release date: 2025-11-20 - Pricing: per-generation — see https://aihubmix.com/model/gemini-3-pro-image-preview ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/edits` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-3-pro-image-preview","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string,null | no | "16:9" | 16:9, 9:16, 1:1, 4:3, 3:4, 3:2, … | | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `image` | - | no | | | | `images` | array,null | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `prompt` | string | yes | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/gemini-3-pro-image-preview/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/gemini-3-pro-image-preview/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/gemini-3-pro-image-preview - Try in browser: https://aihubmix.com/playground?model=gemini-3-pro-image-preview - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-3-pro-image-preview/llms.txt --- # GPT-5.2-Codex Model id on AIHubMix: `gpt-5.2-codex` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-5.2-Codex is an upgraded version of GPT-5.2, optimized for agentic coding tasks in Codex and similar execution environments. The model is specifically enhanced for code generation, modification, refactoring, and automated execution workflows, enabling more efficient participation in multi-step, tool-driven programming processes. GPT-5.2-Codex supports low, medium, high, and xhigh reasoning effort settings, allowing flexible trade-offs between latency, reasoning depth, and token usage depending on task complexity. - Developer: OpenAI - Context window: 400,000 tokens - Input modalities: text, image, PDF - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Release date: 2025-12-11 - Pricing: $1.75/M input tokens, $14/M output tokens, $0.175/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-5.2-codex","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-5.2-codex","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-5.2-codex - Try in browser: https://aihubmix.com/playground?model=gpt-5.2-codex - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-5.2-codex.48c6eaab.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-5.2-codex/llms.txt --- # Doubao Seedream 5.0 Lite Model id on AIHubMix: `doubao-seedream-5.0-lite` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Doubao-Seedream-5.0-lite is the latest image-creation model released by ByteDance. For the first time, the model includes an online retrieval capability, allowing it to integrate real-time web information and improve the timeliness of generated images. At the same time, the model’s intelligence has been further upgraded, enabling it to accurately parse complex instructions and visual content. In addition, the model has been enhanced in terms of breadth of world knowledge, reference consistency, and generation quality in professional scenarios, better meeting enterprise-level visual-creation needs. - Developer: Doubao - Input modalities: text, image - Pricing: per-generation — see https://aihubmix.com/model/doubao-seedream-5.0-lite ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/edits` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"doubao-seedream-5.0-lite","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string,null | no | | 16:9, 9:16, 1:1, 4:3, 3:4, 3:2, … | | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `image` | - | no | | | | `images` | array,null | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `output_format` | string,null | no | "jpeg" | png, jpeg | | `prompt` | string | yes | | | | `response_format` | string,null | no | "url" | url, b64_json, null | | `seed` | integer,null | no | | | | `size` | string,null | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/doubao-seedream-5.0-lite/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/doubao-seedream-5.0-lite/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/doubao-seedream-5.0-lite - Try in browser: https://aihubmix.com/playground?model=doubao-seedream-5.0-lite - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/doubao-seedream-5.0-lite/llms.txt --- # GPT Image 1.5 Model id on AIHubMix: `gpt-image-1.5` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT Image 1.5 is a new image generation model powered by OpenAI’s flagship visual capabilities, comprehensively upgraded for high-quality creative and production workflows. It delivers significant improvements in instruction understanding, fine-grained image editing, and detail preservation, while achieving up to 4× faster generation compared to previous versions — reducing latency without compromising quality. GPT Image 1.5 is well suited for image generation, precise visual editing, and professional content creation, balancing performance with efficiency. - Developer: OpenAI - Input modalities: text, image - Release date: 2025-11-25 - Pricing: per-generation — see https://aihubmix.com/model/gpt-image-1.5 ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/edits` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-image-1.5","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `image` | - | no | | | | `images` | array,null | no | | | | `mask` | - | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `output_format` | string,null | no | "png" | | | `prompt` | string | yes | | | | `size` | string|string|null | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/gpt-image-1.5/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/gpt-image-1.5/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/gpt-image-1.5 - Try in browser: https://aihubmix.com/playground?model=gpt-image-1.5 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-image-1.5/llms.txt --- # GPT 5.2 Model id on AIHubMix: `gpt-5.2` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-5.2 is an advanced general-purpose model that improves on GPT-5.1 with more reliable, flexible, and user-friendly interactions. It delivers clearer responses, stronger instruction-following, and adaptive reasoning that scales from simple requests to complex, multi-step tasks. With enhanced control over tone and structure and support for extended context, GPT-5.2 is well suited for agent workflows, analysis, coding, and cross-domain applications. - Developer: OpenAI - Context window: 400,000 tokens - Input modalities: text, image, PDF - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, File search, Structured outputs, Prompt caching - Release date: 2025-12-11 - Pricing: $1.75/M input tokens, $14/M output tokens, $0.175/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-5.2","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-5.2","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-5.2 - Try in browser: https://aihubmix.com/playground?model=gpt-5.2 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-5.2.8878a4f9.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-5.2/llms.txt --- # GPT 5.2 Chat Model id on AIHubMix: `gpt-5.2-chat-latest` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-5.2Chat refers to the GPT-5.2 snapshot currently used in ChatGPT and is optimized for conversational use cases. While GPT-5.2 is recommended for most API applications, GPT-5.2Chat is ideal for testing the latest improvements in chat-based interactions. - Developer: OpenAI - Context window: 128,000 tokens - Input modalities: text, image - Capabilities: Streaming, Tool calling, Web search, Code interpreter, File search, Structured outputs, Prompt caching - Release date: 2025-12-11 - Pricing: $1.75/M input tokens, $14/M output tokens, $0.175/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-5.2-chat-latest","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-5.2-chat-latest","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-5.2-chat-latest - Try in browser: https://aihubmix.com/playground?model=gpt-5.2-chat-latest - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-5.2-chat-latest.7d59f17c.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-5.2-chat-latest/llms.txt --- # GPT 5.2 High Model id on AIHubMix: `gpt-5.2-high` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-5.2 supports configurable reasoning effort only through the /responses endpoint. To make higher-intensity reasoning available directly via the /chat interface, GPT-5.2-High is provided as a reasoning-enhanced variant of GPT-5.2 with reasoning_effort preset to high. It is designed for tasks that require deeper analysis, stronger result consistency, and greater controllability. By applying more aggressive reasoning strategies and more effective use of extended context, the model delivers clearer and more reliable responses, making it well suited for complex agent workflows, long-chain decision-making, and reliability-critical advanced applications. - Developer: OpenAI - Context window: 400,000 tokens - Input modalities: text, image, PDF - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, File search, Structured outputs, Prompt caching - Pricing: $1.75/M input tokens, $14/M output tokens, $0.175/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-5.2-high","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-5.2-high","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-5.2-high - Try in browser: https://aihubmix.com/playground?model=gpt-5.2-high - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-5.2-high.3a35cb16.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-5.2-high/llms.txt --- # GPT 5.2 Low Model id on AIHubMix: `gpt-5.2-low` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-5.2 supports configuring reasoning strength only through the /responses endpoint. To make lower-overhead reasoning available directly in the /chat endpoint, the GPT-5.2-Low model is provided. This model is based on GPT-5.2 with reasoning_effort preset to low. This model is designed for use cases that are sensitive to response latency and cost. By adopting a lighter reasoning strategy, it delivers stable responses with lower latency and higher throughput. It is well suited for high-concurrency conversations, real-time interactions, basic Q&A, and scenarios where deep reasoning is not required. - Developer: OpenAI - Context window: 400,000 tokens - Input modalities: text, image, PDF - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, File search, Structured outputs, Prompt caching - Pricing: $1.75/M input tokens, $14/M output tokens, $0.175/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-5.2-low","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-5.2-low","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-5.2-low - Try in browser: https://aihubmix.com/playground?model=gpt-5.2-low - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-5.2-low.434ce816.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-5.2-low/llms.txt --- # GPT 5.2 Pro Model id on AIHubMix: `gpt-5.2-pro` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-5.2 pro is available in the Responses API only to enable support for multi-turn model interactions before responding to API requests, and other advanced API features in the future. Since GPT-5.2 pro is designed to tackle tough problems, some requests may take several minutes to finish. To avoid timeout, please set a longer timeout duration. It is recommended to use this under good network conditions. - Developer: OpenAI - Context window: 400,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Web search, File search, Background mode - Release date: 2025-12-11 - Pricing: $21/M input tokens, $168/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-5.2-pro","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-5.2-pro","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-5.2-pro - Try in browser: https://aihubmix.com/playground?model=gpt-5.2-pro - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-5.2-pro.963eff22.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-5.2-pro/llms.txt --- # GPT 5.1 Model id on AIHubMix: `gpt-5.1` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-5 is OpenAI’s most advanced language model, designed for complex tasks that require step-by-step reasoning, precise instruction following, and high reliability. It improves reasoning, code generation, and prompt understanding—including test-time routing and intent cues like “think hard about this”—while reducing hallucination and sycophancy. - Developer: OpenAI - Context window: 400,000 tokens - Input modalities: text, image, PDF - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, File search, Structured outputs, Prompt caching - Release date: 2025-11-13 - Pricing: $1.25/M input tokens, $10/M output tokens, $0.125/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-5.1","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-5.1","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-5.1 - Try in browser: https://aihubmix.com/playground?model=gpt-5.1 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-5.1.d479c0b9.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-5.1/llms.txt --- # GPT-5.1-Codex Max Model id on AIHubMix: `gpt-5.1-codex-max` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-5.1-Codex-Max is a frontier programming model built for the agent-driven era. Powered by an upgraded core reasoning architecture, it is specially trained for complex agentic tasks in software engineering, mathematics, and scientific research. It delivers faster performance, greater stability, and higher token efficiency across the entire development lifecycle, including code generation, refactoring, debugging, and engineering collaboration. With native support for multiple context windows and a built-in compaction mechanism, the model can coherently process millions of tokens within a single task. - Developer: OpenAI - Context window: 400,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Release date: 2025-11-13 - Pricing: $1.25/M input tokens, $10/M output tokens, $0.125/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-5.1-codex-max","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-5.1-codex-max","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-5.1-codex-max - Try in browser: https://aihubmix.com/playground?model=gpt-5.1-codex-max - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-5.1-codex-max.af049d05.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-5.1-codex-max/llms.txt --- # GPT 5.1 Chat Model id on AIHubMix: `gpt-5.1-chat-latest` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-5.1 Chat refers to the GPT-5.1 snapshot currently used in ChatGPT and is optimized for conversational use cases. While GPT-5.1 is recommended for most API applications, GPT-5.1 Chat is ideal for testing the latest improvements in chat-based interactions. - Developer: OpenAI - Context window: 128,000 tokens - Input modalities: text, image - Capabilities: Streaming, Tool calling, Web search, Code interpreter, File search, Structured outputs, Prompt caching - Release date: 2025-11-13 - Pricing: $1.25/M input tokens, $10/M output tokens, $0.125/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-5.1-chat-latest","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-5.1-chat-latest","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-5.1-chat-latest - Try in browser: https://aihubmix.com/playground?model=gpt-5.1-chat-latest - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-5.1-chat-latest.941c6894.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-5.1-chat-latest/llms.txt --- # GPT-5.1-Codex Model id on AIHubMix: `gpt-5.1-codex` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-5.1-Codex is a version of GPT-5 optimized for agentic coding tasks in Codex or similar environments. It's available in the Responses API only and the underlying model snapshot will be regularly updated. - Developer: OpenAI - Context window: 400,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Release date: 2025-11-13 - Pricing: $1.25/M input tokens, $10/M output tokens, $0.125/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-5.1-codex","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-5.1-codex","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-5.1-codex - Try in browser: https://aihubmix.com/playground?model=gpt-5.1-codex - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-5.1-codex.2265cdb9.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-5.1-codex/llms.txt --- # GPT-5.1-Codex Mini Model id on AIHubMix: `gpt-5.1-codex-mini` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-5.1 Codex mini is a smaller, more cost-effective, less-capable version of GPT-5.1-Codex. - Developer: OpenAI - Context window: 400,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Release date: 2025-11-13 - Pricing: $0.25/M input tokens, $2/M output tokens, $0.025/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-5.1-codex-mini","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-5.1-codex-mini","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-5.1-codex-mini - Try in browser: https://aihubmix.com/playground?model=gpt-5.1-codex-mini - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-5.1-codex-mini.40afeded.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-5.1-codex-mini/llms.txt --- # Claude Haiku 4.5 Model id on AIHubMix: `claude-haiku-4-5` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Claude Haiku 4.5 is a fast, affordable, and highly capable AI model, excelling at coding and agentic tasks. Its combination of speed and low cost makes it ideal for powering real-time applications like chatbots, high-volume free services, and specialized "sub-agents" for complex tasks in coding, finance, and research. It can also handle common business tasks like creating office documents and assisting with strategy and analysis. ⚠️ The minimum cache token for claude-haiku-4-5 has been increased from 1,024 to 4,096 tokens. - Developer: Anthropic - Context window: 200,000 tokens - Input modalities: text, image, PDF - Capabilities: Thinking, Streaming, Tool calling, Code interpreter, Computer use, Memory tool, Structured outputs, Citations, Prompt caching - Release date: 2025-10-01 - Pricing: $1.1/M input tokens, $5.5/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"claude-haiku-4-5","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"claude-haiku-4-5","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/claude-haiku-4-5 - Try in browser: https://aihubmix.com/playground?model=claude-haiku-4-5 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/claude-haiku-4-5.b04bf601.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/claude-haiku-4-5/llms.txt --- # Claude Sonnet 4.5 Model id on AIHubMix: `claude-sonnet-4-5` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Sonnet 4.5 is the best model in the world for agents, coding, and computer usage. It is also our most accurate and detailed model for long-running tasks, with enhanced knowledge in coding, finance, and cybersecurity. This model supports a thinking parameter to enable thinking requests in Claude mode. - Developer: Anthropic - Context window: 200,000 tokens - Input modalities: text, image, PDF - Capabilities: Thinking, Streaming, Tool calling, Code interpreter, Computer use, Memory tool, Structured outputs, Citations, Prompt caching - Release date: 2025-09-29 - Pricing: $3.3/M input tokens, $16.5/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"claude-sonnet-4-5","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"claude-sonnet-4-5","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/claude-sonnet-4-5 - Try in browser: https://aihubmix.com/playground?model=claude-sonnet-4-5 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/claude-sonnet-4-5.1b94bc96.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/claude-sonnet-4-5/llms.txt --- # Claude Sonnet 4.5 Thinking Model id on AIHubMix: `claude-sonnet-4-5-think` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Claude Sonnet 4.5 does not enable reasoning mode by default. To access its deep reasoning capabilities, users would typically need to call the native Claude API. To make this capability available through an OpenAI-compatible interface, the claude-sonnet-4-5-think model is provided with reasoning mode pre-enabled and a default 64k-token context window, allowing it to be called directly via the OpenAI unified API. Claude Sonnet 4.5 Think is a reasoning-focused variant of Claude Sonnet 4.5 designed for advanced tasks that require rigorous reasoning, complex decision-making, and long-chain analysis; aside from its enhanced reasoning mechanism, all other capabilities remain consistent with the standard Claude Sonnet 4.5 model, making it well suited for complex problem decomposition, multi-step planning, and logic-intensive analysis. - Developer: Anthropic - Context window: 200,000 tokens - Input modalities: text, image, PDF - Capabilities: Thinking, Streaming, Tool calling, Code interpreter, Computer use, Memory tool, Structured outputs, Citations, Prompt caching - Release date: 2025-09-29 - Pricing: $3.3/M input tokens, $16.5/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"claude-sonnet-4-5-think","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"claude-sonnet-4-5-think","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/claude-sonnet-4-5-think - Try in browser: https://aihubmix.com/playground?model=claude-sonnet-4-5-think - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/claude-sonnet-4-5-think.6004bf2b.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/claude-sonnet-4-5-think/llms.txt --- # Grok 4.20 Multi Agent 0309 Model id on AIHubMix: `grok-4.20-multi-agent-0309` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Grok 4.20 is our newest flagship model with industry-leading speed and agentic tool calling capabilities. It combines the lowest hallucination rate on the market with strict prompt adherance, delivering consistently precise and truthful responses. - Developer: Grok - Context window: 1,000,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Web search, Code interpreter, File search, Structured outputs, Prompt caching, Server-side sessions - Release date: 2026-03-09 - Pricing: $2/M input tokens, $6/M output tokens, $0.2/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"grok-4.20-multi-agent-0309","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"grok-4.20-multi-agent-0309","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/grok-4.20-multi-agent-0309 - Try in browser: https://aihubmix.com/playground?model=grok-4.20-multi-agent-0309 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/grok-4.20-multi-agent-0309.d0634d1e.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/grok-4.20-multi-agent-0309/llms.txt --- # Mistral Large 3 Model id on AIHubMix: `mistral-large-3` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Mistral Large 3 is a MoE model with 67.5B total parameters and 41B active parameters, supporting a 256K-token context window. Trained from scratch on 3,000 NVIDIA H200 GPUs, it is one of the strongest permissively licensed open-weight models available. Designed for advanced reasoning and long-context understanding, Mistral Large 3 delivers performance on par with the best instruction-tuned open-weight models for general-purpose tasks, while also offering image understanding capabilities. Its multilingual strengths are particularly notable for non-English/Chinese languages, making it well-suited for global applications. Typical use cases include enterprise assistants, multilingual customer support, content generation and editing, data analysis over long documents, code assistance, and research workflows that require handling large corpora or complex instructions. With its MoE architecture, Mistral Large 3 balances strong performance with efficient inference, providing a versatile backbone for building reliable, production-grade AI systems. - Developer: Mistral - Context window: 256,000 tokens - Input modalities: text, image - Capabilities: Tool calling, Document input, Structured outputs - Release date: 2024-11-01 - Pricing: $0.5/M input tokens, $1.5/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"mistral-large-3","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"mistral-large-3","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/mistral-large-3 - Try in browser: https://aihubmix.com/playground?model=mistral-large-3 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/mistral-large-3.d704f3a5.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/mistral-large-3/llms.txt --- # CC GLM 5 Model id on AIHubMix: `cc-glm-5` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Supports Claude native interface, can be directly requested in Claude Code. - Developer: Z.AI - Context window: 200,000 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Pricing: $0.06/M input tokens, $0.22/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"cc-glm-5","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"cc-glm-5","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/cc-glm-5 - Try in browser: https://aihubmix.com/playground?model=cc-glm-5 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/cc-glm-5.61927b7a.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/cc-glm-5/llms.txt --- # CC GLM 5 Turbo Model id on AIHubMix: `cc-glm-5-turbo` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Supports Claude native interface, can be directly requested in Claude Code. - Developer: Z.AI - Context window: 204,800 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Pricing: $0.06/M input tokens, $0.22/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"cc-glm-5-turbo","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"cc-glm-5-turbo","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/cc-glm-5-turbo - Try in browser: https://aihubmix.com/playground?model=cc-glm-5-turbo - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/cc-glm-5-turbo.f2f91627.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/cc-glm-5-turbo/llms.txt --- # Cloudflare Glm 5.2 Model id on AIHubMix: `cloudflare-glm-5.2` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms - Developer: Z.AI - Context window: 1,000,000 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Pricing: $1.4/M input tokens, $4.4002/M output tokens, $0.2604/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"cloudflare-glm-5.2","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"cloudflare-glm-5.2","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/cloudflare-glm-5.2 - Try in browser: https://aihubmix.com/playground?model=cloudflare-glm-5.2 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/cloudflare-glm-5.2.93ce1980.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/cloudflare-glm-5.2/llms.txt --- # Gemini 2.5 Flash Image Model id on AIHubMix: `gemini-2.5-flash-image` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Gemini 2.5 Flash Image (Nano-Banana) is a state-of-the-art image generation and editing model that enables seamless blending of multiple images into a single composition while maintaining character consistency for rich visual storytelling. It supports precise, targeted image transformations through natural language instructions and leverages built-in world knowledge for both image generation and editing, making it well suited for creative design, content production, advertising, and visual expression workflows. - Developer: Google - Input modalities: text, image - Release date: 2025-08-26 - Pricing: per-generation — see https://aihubmix.com/model/gemini-2.5-flash-image ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/edits` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-2.5-flash-image","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string|null | no | null | | | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `image` | - | no | | | | `images` | array,null | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `negative_prompt` | string,null | no | | | | `output_format` | string,null | no | "png" | png, jpeg, webp, null | | `prompt` | string | yes | | | | `response_format` | string,null | no | "url" | url, b64_json, null | | `seed` | integer,null | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/gemini-2.5-flash-image/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/gemini-2.5-flash-image/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/gemini-2.5-flash-image - Try in browser: https://aihubmix.com/playground?model=gemini-2.5-flash-image - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-2.5-flash-image/llms.txt --- # grok-4-1-fast-non-reasoning Model id on AIHubMix: `grok-4-1-fast-non-reasoning` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Grok 4.1 is a new conversational model with significant improvements in real-world usability, delivering exceptional performance in creative, emotional, and collaborative interactions. It is more perceptive to nuanced user intent, more engaging to converse with, and more coherent in personality, while fully preserving its core intelligence and reliability. Built on large-scale reinforcement learning infrastructure, the model is optimized for style, personality, helpfulness, and alignment, and leverages frontier agentic reasoning models as reward evaluators to autonomously assess and iterate on responses at scale, significantly enhancing overall interaction quality. - Developer: Grok - Context window: 1,000,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Release date: 2025-11-19 - Pricing: $0.2/M input tokens, $0.5/M output tokens, $0.05/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"grok-4-1-fast-non-reasoning","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"grok-4-1-fast-non-reasoning","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/grok-4-1-fast-non-reasoning - Try in browser: https://aihubmix.com/playground?model=grok-4-1-fast-non-reasoning - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/grok-4-1-fast-non-reasoning.53c5bf21.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/grok-4-1-fast-non-reasoning/llms.txt --- # Grok 4.1 Fast (reasoning) Model id on AIHubMix: `grok-4-1-fast-reasoning` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Grok 4.1 is a new conversational model with significant improvements in real-world usability, delivering exceptional performance in creative, emotional, and collaborative interactions. It is more perceptive to nuanced user intent, more engaging to converse with, and more coherent in personality, while fully preserving its core intelligence and reliability. Built on large-scale reinforcement learning infrastructure, the model is optimized for style, personality, helpfulness, and alignment, and leverages frontier agentic reasoning models as reward evaluators to autonomously assess and iterate on responses at scale, significantly enhancing overall interaction quality. - Developer: Grok - Context window: 1,000,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Release date: 2025-11-19 - Pricing: $0.2/M input tokens, $0.5/M output tokens, $0.05/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"grok-4-1-fast-reasoning","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"grok-4-1-fast-reasoning","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/grok-4-1-fast-reasoning - Try in browser: https://aihubmix.com/playground?model=grok-4-1-fast-reasoning - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/grok-4-1-fast-reasoning.f91dfad0.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/grok-4-1-fast-reasoning/llms.txt --- # Grok Code Fast 1 Model id on AIHubMix: `grok-code-fast-1` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Grok 4.1 is a new conversational model with significant improvements in real-world usability, delivering exceptional performance in creative, emotional, and collaborative interactions. It is more perceptive to nuanced user intent, more engaging to converse with, and more coherent in personality, while fully preserving its core intelligence and reliability. Built on large-scale reinforcement learning infrastructure, the model is optimized for style, personality, helpfulness, and alignment, and leverages frontier agentic reasoning models as reward evaluators to autonomously assess and iterate on responses at scale, significantly enhancing overall interaction quality. - Developer: Grok - Context window: 256,000 tokens - Input modalities: text, image - Capabilities: Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2024-08-25 - Pricing: $1/M input tokens, $2/M output tokens, $0.2/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"grok-code-fast-1","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"grok-code-fast-1","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/grok-code-fast-1 - Try in browser: https://aihubmix.com/playground?model=grok-code-fast-1 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/grok-code-fast-1.9a8ce638.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/grok-code-fast-1/llms.txt --- # K2.6 Code Preview (free) Model id on AIHubMix: `k2.6-code-preview-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > kimi-for-coding-free is a free and open version offered by AIHubMix specifically for Kimi users. To maintain stable service operations, the following usage limits apply: a maximum of 5 requests per minute 500 total requests per day, and a daily quota of 1 million tokens. - Developer: Moonshot AI - Context window: 262,144 tokens - Input modalities: text, image, video - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"k2.6-code-preview-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"k2.6-code-preview-free","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/k2.6-code-preview-free - Try in browser: https://aihubmix.com/playground?model=k2.6-code-preview-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/k2.6-code-preview-free.bb3f3682.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/k2.6-code-preview-free/llms.txt --- # Musesteamer Air Image Model id on AIHubMix: `musesteamer-air-image` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > musesteamer-air-image is a text-to-image model developed by the Baidu Search team aimed at providing extreme cost-effectiveness. It can quickly generate clear images with coherent actions based on user prompts, making it easy to convert users' descriptions into images. - Developer: Baidu - Pricing: per-generation — see https://aihubmix.com/model/musesteamer-air-image ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"musesteamer-air-image","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `images` | array,null | no | | | | `model` | string | yes | | | | `output_format` | string,null | no | "png" | png, jpeg, webp, null | | `prompt` | string | yes | | | | `response_format` | string,null | no | "url" | url, b64_json, null | | `seed` | integer,null | no | | | | `size` | string,null | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/musesteamer-air-image/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/musesteamer-air-image/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/musesteamer-air-image - Try in browser: https://aihubmix.com/playground?model=musesteamer-air-image - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/musesteamer-air-image/llms.txt --- # Qwen3.6 Plus Preview (free) Model id on AIHubMix: `qwen3.6-plus-preview-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > This model has been removed from the platform. - Developer: Qwen - Context window: 1,000,000 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Vision, Video input, Structured outputs, Prompt caching - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen3.6-plus-preview-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"qwen3.6-plus-preview-free","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/qwen3.6-plus-preview-free - Try in browser: https://aihubmix.com/playground?model=qwen3.6-plus-preview-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/qwen3.6-plus-preview-free.590bab70.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen3.6-plus-preview-free/llms.txt --- # Zai Glm 5 Turbo Model id on AIHubMix: `zai-glm-5-turbo` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms - Developer: Z.AI - Context window: 204,800 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Pricing: $1.2/M input tokens, $3.9996/M output tokens, $0.24/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"zai-glm-5-turbo","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"zai-glm-5-turbo","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/zai-glm-5-turbo - Try in browser: https://aihubmix.com/playground?model=zai-glm-5-turbo - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/zai-glm-5-turbo.45297849.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/zai-glm-5-turbo/llms.txt --- # GPT 5 Model id on AIHubMix: `gpt-5` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-5 is OpenAI’s most advanced general-purpose model, delivering major improvements in reasoning, code quality, and overall user experience. It is optimized for complex tasks that require step-by-step reasoning, precise instruction following, and high accuracy in high-stakes scenarios. The model supports test-time routing and advanced prompt understanding, including user-specified intent such as “think hard about this,” while significantly reducing hallucination and sycophancy and improving performance in coding, writing, and health-related tasks. - Developer: OpenAI - Context window: 400,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, File search, Structured outputs, Prompt caching - Release date: 2025-08-07 - Pricing: $1.25/M input tokens, $10/M output tokens, $0.125/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-5","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-5","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-5 - Try in browser: https://aihubmix.com/playground?model=gpt-5 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-5.e4aa7e4c.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-5/llms.txt --- # DeepSeek V3.2 Model id on AIHubMix: `deepseek-v3.2` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > DeepSeek-V3.2 is an efficient large language model equipped with DeepSeek Sparse Attention and reinforced reasoning performance, but its core strength lies in powerful agentic capabilities—enabled by large-scale task-synthesis that tightly integrates reasoning with real-world tool use, delivering robust, compliant, and generalizable agent behaviour. Users can toggle deeper reasoning through the reasoning_enabled switch. - Developer: DeepSeek - Context window: 163,840 tokens - Input modalities: text - Capabilities: Thinking, Tool calling, Structured outputs - Release date: 2025-12-01 - Pricing: $0.58/M input tokens, $1.68/M output tokens, $1/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"deepseek-v3.2","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"deepseek-v3.2","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/deepseek-v3.2 - Try in browser: https://aihubmix.com/playground?model=deepseek-v3.2 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/deepseek-v3.2.084710fa.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/deepseek-v3.2/llms.txt --- # GPT-5-Codex Model id on AIHubMix: `gpt-5-codex` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-5-Codex is a version of GPT-5 optimized for autonomous coding tasks in Codex or similar environments. It is only available in the Responses API, and the underlying model snapshots will be updated regularly. https://docs.aihubmix.com/en/api/Responses-API You can also use it in codex-cll; see https://docs.aihubmix.com/en/api/Codex-CLI for using codex-cll through Aihubmix. - Developer: OpenAI - Context window: 400,000 tokens - Input modalities: text, image - Capabilities: Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Release date: 2025-09-15 - Pricing: $1.25/M input tokens, $10/M output tokens, $0.125/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-5-codex","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-5-codex","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-5-codex - Try in browser: https://aihubmix.com/playground?model=gpt-5-codex - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-5-codex.41eb1abc.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-5-codex/llms.txt --- # DeepSeek V3.1 Thinking Model id on AIHubMix: `DeepSeek-V3.1-Think` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Thinking mode of DeepSeek-V3.1; DeepSeek V3.1 is a text generation model provided by DeepSeek, featuring a hybrid reasoning architecture that achieves an effective integration of thinking and non-thinking modes. - Developer: DeepSeek - Context window: 128,000 tokens - Input modalities: text - Capabilities: Thinking - Release date: 2025-08-21 - Pricing: $0.56/M input tokens, $1.68/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"DeepSeek-V3.1-Think","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"DeepSeek-V3.1-Think","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/DeepSeek-V3.1-Think - Try in browser: https://aihubmix.com/playground?model=DeepSeek-V3.1-Think - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/DeepSeek-V3.1-Think.d7d18105.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/DeepSeek-V3.1-Think/llms.txt --- # GPT 5 Pro Model id on AIHubMix: `gpt-5-pro` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-5 pro uses more compute to think harder and provide consistently better answers. GPT-5 pro is available in the Responses API only to enable support for multi-turn model interactions before responding to API requests, and other advanced API features in the future. Since GPT-5 pro is designed to tackle tough problems, some requests may take several minutes to finish. To avoid timeouts, try using background mode. As our most advanced reasoning model, GPT-5 pro defaults to (and only supports) reasoning.effort: high. GPT-5 pro does not support code interpreter. - Developer: OpenAI - Context window: 400,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Web search, File search, Structured outputs, Background mode - Release date: 2025-10-06 - Pricing: $15/M input tokens, $120/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-5-pro","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-5-pro","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-5-pro - Try in browser: https://aihubmix.com/playground?model=gpt-5-pro - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-5-pro.295c04b0.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-5-pro/llms.txt --- # GPT 5 Mini Model id on AIHubMix: `gpt-5-mini` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-5 mini is a faster, more cost-efficient version of GPT-5. It's great for well-defined tasks and precise prompts. - Developer: OpenAI - Context window: 400,000 tokens - Input modalities: text, image - Capabilities: Streaming, Tool calling, Web search, Code interpreter, File search, Structured outputs - Release date: 2025-08-07 - Pricing: $0.25/M input tokens, $2/M output tokens, $0.025/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-5-mini","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-5-mini","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-5-mini - Try in browser: https://aihubmix.com/playground?model=gpt-5-mini - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-5-mini.ef43ed52.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-5-mini/llms.txt --- # GPT 5 Nano Model id on AIHubMix: `gpt-5-nano` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-5-Nano is the smallest and fastest variant in the GPT-5 system, designed specifically for developer tools and environments that demand rapid interactions and ultra-low latency. While it offers a more lightweight solution with limited reasoning depth compared to its larger counterparts, GPT-5-Nano excels in core capabilities such as instruction-following and maintaining critical safety features. As the successor to GPT-4.1-nano, it provides an optimal choice for cost-sensitive or real-time applications, where efficiency and speed are paramount. Particularly well-suited for summarization and classification tasks, GPT-5-Nano is a powerful tool for developers needing a swift, reliable AI model for streamlined processes. - Developer: OpenAI - Context window: 400,000 tokens - Input modalities: text, image - Capabilities: Streaming, Tool calling, Web search, Code interpreter, File search, Structured outputs, Prompt caching - Release date: 2025-08-07 - Pricing: $0.05/M input tokens, $0.4/M output tokens, $0.005/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-5-nano","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-5-nano","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-5-nano - Try in browser: https://aihubmix.com/playground?model=gpt-5-nano - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-5-nano.23a1a545.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-5-nano/llms.txt --- # GPT 5 Chat Model id on AIHubMix: `gpt-5-chat-latest` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-5 Chat points to the GPT-5 snapshot currently used in ChatGPT. GPT-5 is our next-generation, high-intelligence flagship model. It accepts both text and image inputs, and produces text outputs. - Developer: OpenAI - Context window: 128,000 tokens - Input modalities: text, image - Capabilities: Streaming, Tool calling, Web search, Code interpreter, File search, Structured outputs - Release date: 2025-08-07 - Pricing: $1.25/M input tokens, $10/M output tokens, $0.125/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-5-chat-latest","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-5-chat-latest","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-5-chat-latest - Try in browser: https://aihubmix.com/playground?model=gpt-5-chat-latest - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-5-chat-latest.495fd893.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-5-chat-latest/llms.txt --- # Claude Opus 4.1 Model id on AIHubMix: `claude-opus-4-1` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Opus 4.1 is an upgraded version of Claude Opus 4, with improvements mainly in agent tasks, practical coding, and reasoning. Compared to Opus 4, there is a slight improvement in software engineering accuracy; Opus 4.1 has higher accuracy at 74.5%. - Developer: Anthropic - Context window: 200,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Computer use, Memory tool, Prompt caching - Release date: 2025-08-05 - Pricing: $16.5/M input tokens, $82.5/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"claude-opus-4-1","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"claude-opus-4-1","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/claude-opus-4-1 - Try in browser: https://aihubmix.com/playground?model=claude-opus-4-1 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/claude-opus-4-1.9abcc4eb.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/claude-opus-4-1/llms.txt --- # Kimi K2.5 Model id on AIHubMix: `kimi-k2.5` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Kimi K2.5 is the smartest model of Kimi to date, achieving open-source state-of-the-art (SoTA) performance in Agent, coding, visual understanding, and a series of general intelligent tasks. At the same time, Kimi K2.5 is also the most versatile model of Kimi so far, with a native multimodal architecture design that supports both visual and text input, thinking and non-thinking modes, as well as dialogue and Agent tasks. - Developer: Moonshot AI - Context window: 262,144 tokens - Input modalities: text, image, video - Release date: 2026-01 - Pricing: $0.6/M input tokens, $3/M output tokens, $0.105/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"kimi-k2.5","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"kimi-k2.5","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/kimi-k2.5 - Try in browser: https://aihubmix.com/playground?model=kimi-k2.5 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/kimi-k2.5.7ad386ab.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/kimi-k2.5/llms.txt --- # Qwen3 Max 2026 01-23 Model id on AIHubMix: `qwen3-max-2026-01-23` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > The snapshot version of the Tongyi Qianwen 3 series Max model is from January 23, 2026. By default, it does not require thinking, but thinking mode can be enabled through the enable_thinking parameter, as detailed in the code example. (After enabling thinking by passing parameters, it becomes: Qwen3-Max-Thinking). This model has a total parameter count exceeding one trillion (1T) and a pre-training data volume of up to 36T Tokens, making it the largest and most powerful reasoning model from Alibaba to date. - Developer: Qwen - Context window: 262,144 tokens - Input modalities: text - Release date: 2026-01-23 - Pricing: $0.4508/M input tokens, $1.8032/M output tokens, $0.0902/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen3-max-2026-01-23","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"qwen3-max-2026-01-23","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/qwen3-max-2026-01-23 - Try in browser: https://aihubmix.com/playground?model=qwen3-max-2026-01-23 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/qwen3-max-2026-01-23.239c3bd6.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen3-max-2026-01-23/llms.txt --- # Qwen3 VL Plus Model id on AIHubMix: `qwen3-vl-plus` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > The Qwen3 series visual understanding model achieves an effective fusion of thinking and non-thinking modes. Its visual agent capabilities reach world-class levels on public test sets such as OS World. This version features comprehensive upgrades in visual coding, spatial perception, and multimodal reasoning; visual perception and recognition abilities are greatly enhanced, supporting ultra-long video understanding. - Developer: Qwen - Context window: 262,144 tokens - Input modalities: text, image, video - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2025-09-23 - Pricing: $0.137/M input tokens, $1.37/M output tokens, $0.0274/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen3-vl-plus","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"qwen3-vl-plus","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/qwen3-vl-plus - Try in browser: https://aihubmix.com/playground?model=qwen3-vl-plus - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/qwen3-vl-plus.88afc2fc.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen3-vl-plus/llms.txt --- # CC MiniMax M2.7 Model id on AIHubMix: `cc-minimax-m2.7` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > For Claude Code only - Developer: Minimax - Context window: 204,800 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Prompt caching - Pricing: $0.1/M input tokens, $0.1/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"cc-minimax-m2.7","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"cc-minimax-m2.7","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/cc-minimax-m2.7 - Try in browser: https://aihubmix.com/playground?model=cc-minimax-m2.7 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/cc-minimax-m2.7.bead8c3c.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/cc-minimax-m2.7/llms.txt --- # CC MiniMax M2.7 Highspeed Model id on AIHubMix: `cc-minimax-m2.7-highspeed` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > For Claude Code only - Developer: Minimax - Context window: 204,800 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Prompt caching - Pricing: $0.1/M input tokens, $0.1/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"cc-minimax-m2.7-highspeed","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"cc-minimax-m2.7-highspeed","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/cc-minimax-m2.7-highspeed - Try in browser: https://aihubmix.com/playground?model=cc-minimax-m2.7-highspeed - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/cc-minimax-m2.7-highspeed.93136411.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/cc-minimax-m2.7-highspeed/llms.txt --- # MiniMax M2.5 Model id on AIHubMix: `minimax-m2.5` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > The MiniMax M2.5 is a flagship programming model built for real-world productivity. As a production-grade model natively designed for Agent scenarios, it has achieved state-of-the-art (SOTA) performance in coding, agentic tool use, search, and office work. - Developer: Minimax - Context window: 204,800 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Prompt caching - Release date: 2026-02-12 - Pricing: $0.288/M input tokens, $1.152/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"minimax-m2.5","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"minimax-m2.5","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/minimax-m2.5 - Try in browser: https://aihubmix.com/playground?model=minimax-m2.5 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/minimax-m2.5.8d98e4b8.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/minimax-m2.5/llms.txt --- # MiniMax M2.5 Highspeed Model id on AIHubMix: `minimax-m2.5-highspeed` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > • Same performance as minimax-m2.5 • Significantly faster inference - Developer: Minimax - Context window: 204,800 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Prompt caching - Release date: 2026-02-13 - Pricing: $0.288/M input tokens, $1.152/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"minimax-m2.5-highspeed","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"minimax-m2.5-highspeed","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/minimax-m2.5-highspeed - Try in browser: https://aihubmix.com/playground?model=minimax-m2.5-highspeed - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/minimax-m2.5-highspeed.f5d3bb32.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/minimax-m2.5-highspeed/llms.txt --- # Mm Minimax M2.7 Highspeed Model id on AIHubMix: `mm-minimax-m2.7-highspeed` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > For Claude Code only - Developer: Minimax - Context window: 204,800 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Prompt caching - Pricing: $0.1/M input tokens, $0.1/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"mm-minimax-m2.7-highspeed","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"mm-minimax-m2.7-highspeed","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/mm-minimax-m2.7-highspeed - Try in browser: https://aihubmix.com/playground?model=mm-minimax-m2.7-highspeed - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/mm-minimax-m2.7-highspeed.c96980de.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/mm-minimax-m2.7-highspeed/llms.txt --- # Coding MiniMax M2.7 Model id on AIHubMix: `coding-minimax-m2.7` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms - Developer: Minimax - Context window: 204,800 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Prompt caching - Release date: 2026-03-18 - Pricing: $0.2/M input tokens, $0.2/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"coding-minimax-m2.7","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"coding-minimax-m2.7","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/coding-minimax-m2.7 - Try in browser: https://aihubmix.com/playground?model=coding-minimax-m2.7 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/coding-minimax-m2.7.4db66070.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/coding-minimax-m2.7/llms.txt --- # Coding MiniMax M2.7 Highspeed Model id on AIHubMix: `coding-minimax-m2.7-highspeed` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms - Developer: Minimax - Context window: 204,800 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Prompt caching - Release date: 2026-03-18 - Pricing: $0.2/M input tokens, $0.2/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"coding-minimax-m2.7-highspeed","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"coding-minimax-m2.7-highspeed","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/coding-minimax-m2.7-highspeed - Try in browser: https://aihubmix.com/playground?model=coding-minimax-m2.7-highspeed - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/coding-minimax-m2.7-highspeed.84062969.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/coding-minimax-m2.7-highspeed/llms.txt --- # CC MiniMax M2.5 Model id on AIHubMix: `cc-minimax-m2.5` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > For Claude Code only - Developer: Minimax - Context window: 204,800 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Prompt caching - Pricing: $0.1/M input tokens, $0.1/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"cc-minimax-m2.5","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"cc-minimax-m2.5","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/cc-minimax-m2.5 - Try in browser: https://aihubmix.com/playground?model=cc-minimax-m2.5 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/cc-minimax-m2.5.12777d66.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/cc-minimax-m2.5/llms.txt --- # CC MiniMax M2.5 Highspeed Model id on AIHubMix: `cc-minimax-m2.5-highspeed` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > For Claude Code only - Developer: Minimax - Context window: 204,800 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Prompt caching - Pricing: $0.1/M input tokens, $0.1/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"cc-minimax-m2.5-highspeed","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"cc-minimax-m2.5-highspeed","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/cc-minimax-m2.5-highspeed - Try in browser: https://aihubmix.com/playground?model=cc-minimax-m2.5-highspeed - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/cc-minimax-m2.5-highspeed.297c4ff3.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/cc-minimax-m2.5-highspeed/llms.txt --- # Coding MiniMax M2.5 Model id on AIHubMix: `coding-minimax-m2.5` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms - Developer: Minimax - Context window: 204,800 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Prompt caching - Release date: 2026-02-12 - Pricing: $0.2/M input tokens, $0.2/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"coding-minimax-m2.5","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"coding-minimax-m2.5","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/coding-minimax-m2.5 - Try in browser: https://aihubmix.com/playground?model=coding-minimax-m2.5 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/coding-minimax-m2.5.eba3cac5.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/coding-minimax-m2.5/llms.txt --- # Coding MiniMax M2.5 Highspeed Model id on AIHubMix: `coding-minimax-m2.5-highspeed` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms - Developer: Minimax - Context window: 204,800 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Prompt caching - Release date: 2026-02-13 - Pricing: $0.2/M input tokens, $0.2/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"coding-minimax-m2.5-highspeed","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"coding-minimax-m2.5-highspeed","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/coding-minimax-m2.5-highspeed - Try in browser: https://aihubmix.com/playground?model=coding-minimax-m2.5-highspeed - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/coding-minimax-m2.5-highspeed.3f9e6643.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/coding-minimax-m2.5-highspeed/llms.txt --- # Doubao Seedream 4.5 Model id on AIHubMix: `doubao-seedream-4-5` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Seedream 4.5 is ByteDance's latest multimodal image model, integrating capabilities such as text-to-image, image-to-image, and multi-image output, along with incorporating common sense and reasoning abilities. Compared to the previous 4.0 model, it significantly improves generation quality, offering better editing consistency and multi-image fusion effects, with more precise control over image details. The generation of small text and small faces is more natural. - Developer: ByteDance - Input modalities: text, image - Pricing: per-generation — see https://aihubmix.com/model/doubao-seedream-4-5 ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/edits` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"doubao-seedream-4-5","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string,null | no | | 16:9, 9:16, 1:1, 4:3, 3:4, 3:2, … | | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `image` | - | no | | | | `images` | array,null | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `prompt` | string | yes | | | | `response_format` | string,null | no | "url" | url, b64_json, null | | `seed` | integer,null | no | | | | `size` | string,null | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/doubao-seedream-4-5/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/doubao-seedream-4-5/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/doubao-seedream-4-5 - Try in browser: https://aihubmix.com/playground?model=doubao-seedream-4-5 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/doubao-seedream-4-5/llms.txt --- # Sora 2 Model id on AIHubMix: `sora-2` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Sora-2 is the next-generation text-to-video model evolved from Sora, optimized for higher visual realism, stronger physical consistency, and longer temporal coherence. It delivers more stable character consistency, complex motion rendering, camera control, and narrative continuity, while supporting higher resolutions and minute-level video generation for film production, advertising, virtual content creation, and creative multimedia workflows. - Developer: OpenAI - Release date: 2025-10-06 - Pricing: per-generation — see https://aihubmix.com/model/sora-2 ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /ai/v1/videos/{id}` until a terminal state - `POST /v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /v1/videos/{id}` until a terminal state ## Example ```bash curl -s https://aihubmix.com/ai/v1/videos \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"sora-2","prompt":""}' ``` The response contains a task id — poll `GET https://aihubmix.com/ai/v1/videos/` until `status` is terminal, then read the output URL(s). Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (video endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `duration` | integer|null | no | 4 | | | `extra` | object,null | no | | | | `frame_images` | array,null | no | | | | `generate_audio` | boolean,null | no | | | | `input_references` | array,null | no | | | | `model` | string | yes | | | | `prompt` | string | yes | | | | `size` | string,null | no | | | | `webhook_url` | string,null | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/sora-2/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/sora-2/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/sora-2 - Try in browser: https://aihubmix.com/playground?model=sora-2 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/sora-2/llms.txt --- # Sora 2 Pro Model id on AIHubMix: `sora-2-pro` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > OpenAI video model Sora2-pro official API. - Developer: OpenAI - Release date: 2025-10-06 - Pricing: per-generation — see https://aihubmix.com/model/sora-2-pro ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /ai/v1/videos/{id}` until a terminal state - `POST /v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /v1/videos/{id}` until a terminal state ## Example ```bash curl -s https://aihubmix.com/ai/v1/videos \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"sora-2-pro","prompt":""}' ``` The response contains a task id — poll `GET https://aihubmix.com/ai/v1/videos/` until `status` is terminal, then read the output URL(s). Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (video endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `duration` | integer|null | no | 4 | | | `extra` | object,null | no | | | | `frame_images` | array,null | no | | | | `generate_audio` | boolean,null | no | | | | `input_references` | array,null | no | | | | `model` | string | yes | | | | `prompt` | string | yes | | | | `size` | string,null | no | | | | `webhook_url` | string,null | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/sora-2-pro/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/sora-2-pro/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/sora-2-pro - Try in browser: https://aihubmix.com/playground?model=sora-2-pro - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/sora-2-pro/llms.txt --- # CC MiniMax M2.1 Model id on AIHubMix: `cc-minimax-m2.1` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > For Claude Code only - Developer: Minimax - Context window: 204,800 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Prompt caching - Pricing: $0.1/M input tokens, $0.1/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"cc-minimax-m2.1","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"cc-minimax-m2.1","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/cc-minimax-m2.1 - Try in browser: https://aihubmix.com/playground?model=cc-minimax-m2.1 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/cc-minimax-m2.1.420cdbd2.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/cc-minimax-m2.1/llms.txt --- # Coding GLM 4.7 Model id on AIHubMix: `coding-glm-4.7` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Only supports OpenAI-compatible formats. - Developer: Z.AI - Context window: 200,000 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2025-12-22 - Pricing: $0.06/M input tokens, $0.22/M output tokens, $0.011/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"coding-glm-4.7","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"coding-glm-4.7","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/coding-glm-4.7 - Try in browser: https://aihubmix.com/playground?model=coding-glm-4.7 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/coding-glm-4.7.947d0813.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/coding-glm-4.7/llms.txt --- # Coding MiniMax M2.1 Model id on AIHubMix: `coding-minimax-m2.1` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms - Developer: Minimax - Context window: 204,800 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Prompt caching - Release date: 2025-12-23 - Pricing: $0.2/M input tokens, $0.2/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"coding-minimax-m2.1","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"coding-minimax-m2.1","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/coding-minimax-m2.1 - Try in browser: https://aihubmix.com/playground?model=coding-minimax-m2.1 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/coding-minimax-m2.1.e3e82d39.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/coding-minimax-m2.1/llms.txt --- # Coding MiniMax M2.1 (free) Model id on AIHubMix: `coding-minimax-m2.1-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > coding-minimax-m2.1-free is a free and open version offered by AIHubMix specifically for MiniMax users. To maintain stable service operations, the following usage limits apply: a maximum of 5 requests per minute, 500 total requests per day, and a daily quota of 1 million tokens. - Developer: Minimax - Context window: 204,800 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Prompt caching - Release date: 2025-12-23 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"coding-minimax-m2.1-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"coding-minimax-m2.1-free","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/coding-minimax-m2.1-free - Try in browser: https://aihubmix.com/playground?model=coding-minimax-m2.1-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/coding-minimax-m2.1-free.8327f163.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/coding-minimax-m2.1-free/llms.txt --- # GPT 4o Mini Audio Preview Model id on AIHubMix: `gpt-4o-mini-audio-preview` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > openai声音输入输出模型,价格和官方一致,暂时只展示文字部分价格,声音价格见openai官网;后台扣费和官方一致 - Developer: OpenAI - Context window: 128,000 tokens - Input modalities: text, audio - Capabilities: Streaming, Tool calling, Audio output, Multimodal output - Pricing: $0.15/M input tokens, $0.6/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-4o-mini-audio-preview","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-4o-mini-audio-preview","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-4o-mini-audio-preview - Try in browser: https://aihubmix.com/playground?model=gpt-4o-mini-audio-preview - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-4o-mini-audio-preview.89f553f3.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-4o-mini-audio-preview/llms.txt --- # MiniMax M2.1 Model id on AIHubMix: `minimax-m2.1` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > MiniMax-M2.1 redefines efficiency for intelligent agents. It is a compact, fast, and cost-effective MoE model with a total of 230 billion parameters and 10 billion active parameters, designed for top performance in coding and intelligent agent tasks while maintaining strong general intelligence. With only 10 billion active parameters, MiniMax-M2 delivers the complex end-to-end tool usage performance expected from today's leading models, but in a more streamlined form factor, making deployment and scaling easier than ever before. - Developer: Minimax - Context window: 204,800 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Prompt caching - Release date: 2025-12-23 - Pricing: $0.288/M input tokens, $1.152/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"minimax-m2.1","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"minimax-m2.1","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/minimax-m2.1 - Try in browser: https://aihubmix.com/playground?model=minimax-m2.1 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/minimax-m2.1.68f073b4.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/minimax-m2.1/llms.txt --- # O3 Model id on AIHubMix: `o3` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > OpenAI o3 is a powerful model across multiple domains, setting a new standard for coding, math, science, and visual reasoning tasks. - Developer: OpenAI - Context window: 200,000 tokens - Input modalities: text, image, PDF - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, File search, Structured outputs, Prompt caching - Release date: 2025-04-16 - Pricing: $2/M input tokens, $8/M output tokens, $0.5/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"o3","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"o3","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/o3 - Try in browser: https://aihubmix.com/playground?model=o3 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/o3.b8231a16.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/o3/llms.txt --- # Wan2.6 I2v Model id on AIHubMix: `wan2.6-i2v` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Wan 2.6 - Text-to-Video generation features intelligent storyboard scheduling supporting multi-shot narration, higher quality sound generation, stable multi-person dialogue, more natural and realistic voice tones, and supports video generation up to 15 seconds in length. - Developer: Qwen - Input modalities: image, text - Pricing: per-generation — see https://aihubmix.com/model/wan2.6-i2v ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /ai/v1/videos/{id}` until a terminal state - `POST /v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /v1/videos/{id}` until a terminal state ## Example ```bash curl -s https://aihubmix.com/ai/v1/videos \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"wan2.6-i2v","prompt":""}' ``` The response contains a task id — poll `GET https://aihubmix.com/ai/v1/videos/` until `status` is terminal, then read the output URL(s). Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (video endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `duration` | integer|null | no | 5 | | | `extra` | object,null | no | | | | `frame_images` | array,null | no | | | | `model` | string | yes | | | | `prompt` | string | yes | | | | `resolution` | string,null | no | "1080p" | 720p, 1080p, null | | `seed` | integer,null | no | | | | `webhook_url` | string,null | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/wan2.6-i2v/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/wan2.6-i2v/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/wan2.6-i2v - Try in browser: https://aihubmix.com/playground?model=wan2.6-i2v - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/wan2.6-i2v/llms.txt --- # Wan2.6 T2v Model id on AIHubMix: `wan2.6-t2v` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Wan 2.6 - Text-to-Video generation features intelligent storyboard scheduling supporting multi-shot narration, higher quality sound generation, stable multi-person dialogue, more natural and realistic voice tones, and supports video generation up to 15 seconds in length. - Developer: Qwen - Input modalities: text - Pricing: per-generation — see https://aihubmix.com/model/wan2.6-t2v ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /ai/v1/videos/{id}` until a terminal state - `POST /v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /v1/videos/{id}` until a terminal state ## Example ```bash curl -s https://aihubmix.com/ai/v1/videos \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"wan2.6-t2v","prompt":""}' ``` The response contains a task id — poll `GET https://aihubmix.com/ai/v1/videos/` until `status` is terminal, then read the output URL(s). Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (video endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `duration` | integer|null | no | 5 | | | `extra` | object,null | no | | | | `model` | string | yes | | | | `prompt` | string | yes | | | | `seed` | integer,null | no | | | | `size` | string,null | no | | | | `webhook_url` | string,null | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/wan2.6-t2v/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/wan2.6-t2v/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/wan2.6-t2v - Try in browser: https://aihubmix.com/playground?model=wan2.6-t2v - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/wan2.6-t2v/llms.txt --- # Coding GLM 4.6 Model id on AIHubMix: `coding-glm-4.6` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms - Developer: Z.AI - Context window: 200,000 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2025-09-30 - Pricing: $0.06/M input tokens, $0.22/M output tokens, $0.011/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"coding-glm-4.6","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"coding-glm-4.6","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/coding-glm-4.6 - Try in browser: https://aihubmix.com/playground?model=coding-glm-4.6 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/coding-glm-4.6.14256f3e.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/coding-glm-4.6/llms.txt --- # Coding GLM 4.6 (free) Model id on AIHubMix: `coding-glm-4.6-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > coding-glm-4.6-free is the open and free version of coding-glm-4.6. To ensure stable service performance, usage limits are in place: up to 5 requests per minute, 500 requests per day, and a daily token allowance of 1 million. - Developer: Z.AI - Context window: 200,000 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2025-09-30 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"coding-glm-4.6-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"coding-glm-4.6-free","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/coding-glm-4.6-free - Try in browser: https://aihubmix.com/playground?model=coding-glm-4.6-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/coding-glm-4.6-free.c078df71.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/coding-glm-4.6-free/llms.txt --- # Coding MiniMax M2 Model id on AIHubMix: `coding-minimax-m2` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > coding-minimax-m2 is a free and open version offered by AIHubMix specifically for MiniMax users. To maintain stable service operations, the following usage limits apply: a maximum of 10 requests per minute, 1,000 total requests per day, and a daily quota of 5 million tokens.204800 - Developer: Minimax - Context window: 204,800 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Prompt caching - Release date: 2025-10-27 - Pricing: $0.2/M input tokens, $0.2/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"coding-minimax-m2","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"coding-minimax-m2","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/coding-minimax-m2 - Try in browser: https://aihubmix.com/playground?model=coding-minimax-m2 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/coding-minimax-m2.0517a9b4.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/coding-minimax-m2/llms.txt --- # Coding MiniMax M2 (free) Model id on AIHubMix: `coding-minimax-m2-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > coding-minimax-m2-free is a free and open version offered by AIHubMix specifically for MiniMax users. To maintain stable service operations, the following usage limits apply: a maximum of 5 requests per minute, 500 total requests per day, and a daily quota of 1 million tokens. - Developer: Minimax - Context window: 204,800 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Prompt caching - Release date: 2025-10-27 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"coding-minimax-m2-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"coding-minimax-m2-free","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/coding-minimax-m2-free - Try in browser: https://aihubmix.com/playground?model=coding-minimax-m2-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/coding-minimax-m2-free.44977652.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/coding-minimax-m2-free/llms.txt --- # Flux 2 Flex Model id on AIHubMix: `flux-2-flex` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > FLUX.2 is purpose-built for real-world creative production workflows. It delivers high-quality images while maintaining character and style consistency across multiple reference images, shows exceptional understanding and execution of structured prompts, and supports complex text reading and writing. It also adheres to brand guidelines, handles lighting, layout, and logo elements with stability, and enables image editing at resolutions up to 4MP — all while preserving fine details, striking a balance between creativity and professional-grade visual output. - Developer: Flux - Input modalities: text, image - Release date: 2025-11-25 - Pricing: per-generation — see https://aihubmix.com/model/flux-2-flex ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/edits` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"flux-2-flex","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `image` | - | no | | | | `images` | array,null | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `output_format` | string,null | no | "png" | png, jpeg, webp, null | | `prompt` | string | yes | | | | `response_format` | string,null | no | "url" | url, b64_json, null | | `seed` | integer,null | no | | | | `size` | string,null | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/flux-2-flex/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/flux-2-flex/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/flux-2-flex - Try in browser: https://aihubmix.com/playground?model=flux-2-flex - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/flux-2-flex/llms.txt --- # Flux 2 Pro Model id on AIHubMix: `flux-2-pro` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > FLUX.2 is purpose-built for real-world creative production workflows. It delivers high-quality images while maintaining character and style consistency across multiple reference images, shows exceptional understanding and execution of structured prompts, and supports complex text reading and writing. It also adheres to brand guidelines, handles lighting, layout, and logo elements with stability, and enables image editing at resolutions up to 4MP — all while preserving fine details, striking a balance between creativity and professional-grade visual output. - Developer: Flux - Input modalities: text, image - Release date: 2025-11-25 - Pricing: per-generation — see https://aihubmix.com/model/flux-2-pro ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/edits` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"flux-2-pro","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `image` | - | no | | | | `images` | array,null | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `output_format` | string,null | no | "png" | png, jpeg, webp, null | | `prompt` | string | yes | | | | `response_format` | string,null | no | "url" | url, b64_json, null | | `seed` | integer,null | no | | | | `size` | string,null | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/flux-2-pro/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/flux-2-pro/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/flux-2-pro - Try in browser: https://aihubmix.com/playground?model=flux-2-pro - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/flux-2-pro/llms.txt --- # Gemini 2.5 Pro Model id on AIHubMix: `gemini-2.5-pro` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Gemini 2.5 Pro is an advanced reasoning model developed by Google, optimized for solving highly complex problems across multiple domains. It can deeply understand large-scale information from diverse sources, including text, audio, images, video, and even entire codebases. The model demonstrates strong reasoning capabilities in coding, mathematics, and STEM-related tasks, and supports long-context analysis for large datasets, codebases, and technical documentation. - Developer: Google - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio, PDF - Capabilities: Thinking, Tool calling, Web search, URL context, Code interpreter, File search, Structured outputs, Prompt caching - Release date: 2025-06-17 - Pricing: $1.25/M input tokens, $10/M output tokens, $0.125/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-2.5-pro","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gemini-2.5-pro","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gemini-2.5-pro - Try in browser: https://aihubmix.com/playground?model=gemini-2.5-pro - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gemini-2.5-pro.bbb32099.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-2.5-pro/llms.txt --- # GLM 4.6 Model id on AIHubMix: `glm-4.6` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GLM-4.6 is Zhipu’s latest flagship model (total parameters 355B, activation parameters 32B), comprehensively surpassing GLM-4.5. Its coding capability is aligned with Claude Sonnet 4, making it a top domestic coding model; the context window has been expanded from 128K to 200K, better suited for long code and agent tasks; inference capabilities have been significantly enhanced and support tool invocation during processing; improvements have been made in tool calling, search agents, writing style, role play, and multilingual translation. The model is named glm-4.6 and is provided by three vendors, with calls prioritized to the Sophnet platform. - Developer: Z.AI - Context window: 200,000 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2025-09-30 - Pricing: $0.274/M input tokens, $1.0959/M output tokens, $0.0548/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"glm-4.6","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"glm-4.6","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/glm-4.6 - Try in browser: https://aihubmix.com/playground?model=glm-4.6 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/glm-4.6.46ccd36e.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/glm-4.6/llms.txt --- # GLM 4.6 Vision Model id on AIHubMix: `glm-4.6v` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Zhipu's latest visual reasoning model achieves state-of-the-art visual understanding accuracy at the same scale upon release. It natively supports tool invocation, can automatically complete tasks, supports ultra-long 128K context length, and allows flexible toggling of reasoning. - Developer: Z.AI - Context window: 131,072 tokens - Input modalities: text, image, video, document - Capabilities: Thinking, Streaming, Tool calling, Document input, Prompt caching - Release date: 2025-12-08 - Pricing: $0.137/M input tokens, $0.411/M output tokens, $0.0274/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"glm-4.6v","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"glm-4.6v","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/glm-4.6v - Try in browser: https://aihubmix.com/playground?model=glm-4.6v - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/glm-4.6v.46c8fc2f.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/glm-4.6v/llms.txt --- # Kimi For Coding (free) Model id on AIHubMix: `kimi-for-coding-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > kimi-for-coding-free is a free and open version offered by AIHubMix specifically for Kimi users. To maintain stable service operations, the following usage limits apply: a maximum of 5 requests per minute 500 total requests per day, and a daily quota of 1 million tokens. - Developer: Moonshot AI - Context window: 256,000 tokens - Input modalities: text, image, video - Release date: 2026-06-12 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"kimi-for-coding-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"kimi-for-coding-free","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/kimi-for-coding-free - Try in browser: https://aihubmix.com/playground?model=kimi-for-coding-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/kimi-for-coding-free.bd049d97.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/kimi-for-coding-free/llms.txt --- # O3 Pro Model id on AIHubMix: `o3-pro` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > o3-pro This model only supports Requests API interface requests.The model's thinking time is relatively long, so the response will be slow. - Developer: OpenAI - Context window: 200,000 tokens - Input modalities: text, image - Capabilities: Tool calling, Web search, File search, Structured outputs, Background mode - Release date: 2025-06-10 - Pricing: $20/M input tokens, $80/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"o3-pro","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"o3-pro","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/o3-pro - Try in browser: https://aihubmix.com/playground?model=o3-pro - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/o3-pro.a4f8072b.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/o3-pro/llms.txt --- # Step 3.5 Flash Model id on AIHubMix: `step-3.5-flash` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > step-3.5-flash is stepfun's flagship inference model, designed for high-complexity tasks that require deep reasoning and fast execution. It excels at decomposing multi-step problems, performing tool calls, and maintaining consistency across massive datasets. It is the preferred choice for complex workloads such as long-context agents, advanced software engineering, and end-to-end research automation. - Developer: StepFun - Context window: 256,000 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Prompt caching - Release date: 2026-01-29 - Pricing: $0.11/M input tokens, $0.33/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"step-3.5-flash","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"step-3.5-flash","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/step-3.5-flash - Try in browser: https://aihubmix.com/playground?model=step-3.5-flash - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/step-3.5-flash.6559a4bb.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/step-3.5-flash/llms.txt --- # Wan2.5 I2v Preview Model id on AIHubMix: `wan2.5-i2v-preview` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Tongyi Wanxiang 2.5 - Text-to-Video Preview features a newly upgraded technical architecture, supporting sound generation synchronized with visuals, 10-second long video generation, stronger instruction-following capabilities, and further improvements in motion ability and visual quality. - Developer: Qwen - Input modalities: text, image - Pricing: per-generation — see https://aihubmix.com/model/wan2.5-i2v-preview ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /ai/v1/videos/{id}` until a terminal state - `POST /v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /v1/videos/{id}` until a terminal state ## Example ```bash curl -s https://aihubmix.com/ai/v1/videos \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"wan2.5-i2v-preview","prompt":""}' ``` The response contains a task id — poll `GET https://aihubmix.com/ai/v1/videos/` until `status` is terminal, then read the output URL(s). Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (video endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `duration` | integer|null | no | 5 | | | `extra` | object,null | no | | | | `frame_images` | array,null | no | | | | `generate_audio` | boolean,null | no | | | | `model` | string | yes | | | | `prompt` | string | yes | | | | `resolution` | string,null | no | "1080p" | 480p, 720p, 1080p, null | | `seed` | integer,null | no | | | | `webhook_url` | string,null | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/wan2.5-i2v-preview/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/wan2.5-i2v-preview/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/wan2.5-i2v-preview - Try in browser: https://aihubmix.com/playground?model=wan2.5-i2v-preview - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/wan2.5-i2v-preview/llms.txt --- # Wan2.5 T2v Preview Model id on AIHubMix: `wan2.5-t2v-preview` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Tongyi Wanxiang 2.5 - Text-to-Video Preview, newly upgraded model architecture, supports sound generation synchronized with visuals, supports 10-second long video generation, enhanced instruction compliance, improved motion capability, and further enhanced visual quality. - Developer: Qwen - Input modalities: text - Pricing: per-generation — see https://aihubmix.com/model/wan2.5-t2v-preview ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /ai/v1/videos/{id}` until a terminal state - `POST /v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /v1/videos/{id}` until a terminal state ## Example ```bash curl -s https://aihubmix.com/ai/v1/videos \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"wan2.5-t2v-preview","prompt":""}' ``` The response contains a task id — poll `GET https://aihubmix.com/ai/v1/videos/` until `status` is terminal, then read the output URL(s). Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (video endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `duration` | integer|null | no | 5 | | | `extra` | object,null | no | | | | `generate_audio` | boolean,null | no | | | | `model` | string | yes | | | | `prompt` | string | yes | | | | `seed` | integer,null | no | | | | `size` | string,null | no | | | | `webhook_url` | string,null | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/wan2.5-t2v-preview/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/wan2.5-t2v-preview/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/wan2.5-t2v-preview - Try in browser: https://aihubmix.com/playground?model=wan2.5-t2v-preview - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/wan2.5-t2v-preview/llms.txt --- # Gemini 2.5 Pro Search Model id on AIHubMix: `gemini-2.5-pro-search` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > gemini-2.5-pro-search integrates Google's official search functionality; the search feature will have an additional separate fee log directly incorporated into the scoring, with detailed logs not displayed; this will be fixed and displayed later; only supports OpenAI-compatible formats for invocation, does not support Gemini SDK; for Gemini's native SDK, please set parameters directly using the official search parameters. - Developer: Google - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio, PDF - Capabilities: Thinking, Tool calling, Web search, URL context, Code interpreter, File search, Structured outputs, Prompt caching - Release date: 2025-06-17 - Pricing: $1.25/M input tokens, $10/M output tokens, $0.125/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-2.5-pro-search","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gemini-2.5-pro-search","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gemini-2.5-pro-search - Try in browser: https://aihubmix.com/playground?model=gemini-2.5-pro-search - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gemini-2.5-pro-search.c163e84c.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-2.5-pro-search/llms.txt --- # Kimi K2 Thinking Model id on AIHubMix: `kimi-k2-thinking` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Kimi K2 Thinking is Moonshot AI's most advanced open-source inference model to date, extending the K2 series into intelligent agent and long-context inference domains. The model is built on the trillion-parameter mixture of experts (MoE) architecture introduced by Kimi K2, activating 32 billion parameters per forward pass and supporting a context window of 256,000 tokens. - Developer: Moonshot AI - Context window: 262,144 tokens - Input modalities: text - Capabilities: Thinking, Tool calling - Release date: 2025-11-06 - Pricing: $0.548/M input tokens, $2.192/M output tokens, $0.137/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"kimi-k2-thinking","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"kimi-k2-thinking","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/kimi-k2-thinking - Try in browser: https://aihubmix.com/playground?model=kimi-k2-thinking - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/kimi-k2-thinking.aa1951d8.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/kimi-k2-thinking/llms.txt --- # GLM 4.5 Vision Model id on AIHubMix: `glm-4.5v` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GLM-4.5V is a vision-language foundational model designed for multimodal agent applications. Based on a mixture-of-experts (MoE) architecture, it has 106 billion parameters and 12 billion active parameters. It delivers outstanding performance in video understanding, image question answering, OCR, and document parsing, and achieves significant improvements in front-end web encoding, basic reasoning, and spatial reasoning. - Developer: Z.AI - Context window: 65,536 tokens - Input modalities: text, image, video, document - Capabilities: Thinking, Streaming, Document input, Prompt caching - Release date: 2025-08-11 - Pricing: $0.274/M input tokens, $0.822/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"glm-4.5v","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"glm-4.5v","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/glm-4.5v - Try in browser: https://aihubmix.com/playground?model=glm-4.5v - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/glm-4.5v.2072f13b.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/glm-4.5v/llms.txt --- # Gemini 2.5 Flash Search Model id on AIHubMix: `gemini-2.5-flash-search` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > gemini-2.5-flash-search integrates Google's official search functionality; the search feature will have an additional separate fee log directly incorporated into the scoring, with detailed logs not displayed; this will be fixed and displayed later; only supports OpenAI-compatible formats for invocation, does not support Gemini SDK; for Gemini's native SDK, please set parameters directly using the official search parameters. - Developer: Google - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio, PDF - Capabilities: Thinking, Tool calling, Web search, URL context, Code interpreter, File search, Structured outputs, Prompt caching - Release date: 2025-06-17 - Pricing: $0.3/M input tokens, $2.499/M output tokens, $0.03/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-2.5-flash-search","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gemini-2.5-flash-search","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gemini-2.5-flash-search - Try in browser: https://aihubmix.com/playground?model=gemini-2.5-flash-search - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gemini-2.5-flash-search.4d84b825.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-2.5-flash-search/llms.txt --- # Gemini 2.5 Flash Preview 05-20 (no think) Model id on AIHubMix: `gemini-2.5-flash-preview-05-20-nothink` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Gemini-2.5-flash-preview-05-20 is enabled by default for thinking; to disable it, request the name gemini-2.5-flash-preview-05-20-nothink.Only OpenAI-compatible format calls are supported; Gemini SDK is not supported. For the native Gemini SDK, please set the parameter budget=0 directly. - Developer: Google - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio - Pricing: $0.3/M input tokens, $2.499/M output tokens, $0.03/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-2.5-flash-preview-05-20-nothink","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gemini-2.5-flash-preview-05-20-nothink","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gemini-2.5-flash-preview-05-20-nothink - Try in browser: https://aihubmix.com/playground?model=gemini-2.5-flash-preview-05-20-nothink - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gemini-2.5-flash-preview-05-20-nothink.c4919650.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-2.5-flash-preview-05-20-nothink/llms.txt --- # Gemini 2.5 Flash Preview 05-20 Search Model id on AIHubMix: `gemini-2.5-flash-preview-05-20-search` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Gemini-2.5 Flash Preview 05-20 Search integrates Google's official search functionality; the search feature will have an additional separate fee log directly integrated into the scoring deduction, with detailed logs not displayed. It will be fixed and displayed later. Only OpenAI-compatible formats are supported for invocation; Gemini SDK is not supported. For Gemini's native SDK, please set parameters directly using the official search parameters. - Developer: Google - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio - Release date: 2025-05-20 - Pricing: $0.3/M input tokens, $2.499/M output tokens, $0.03/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-2.5-flash-preview-05-20-search","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gemini-2.5-flash-preview-05-20-search","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gemini-2.5-flash-preview-05-20-search - Try in browser: https://aihubmix.com/playground?model=gemini-2.5-flash-preview-05-20-search - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gemini-2.5-flash-preview-05-20-search.02663cf3.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-2.5-flash-preview-05-20-search/llms.txt --- # GPT Image 1 Model id on AIHubMix: `gpt-image-1` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Azure OpenAI’s gpt-image-1 image generation API offers both text-to-image generation and image-to-image editing with text guidance capabilities. Before using this API, please ensure you have the latest OpenAI package installed by running pip install -U openai. - Developer: OpenAI - Input modalities: text, image - Release date: 2025-04-24 - Pricing: per-generation — see https://aihubmix.com/model/gpt-image-1 ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/edits` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-image-1","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `image` | - | no | | | | `images` | array,null | no | | | | `mask` | - | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `output_format` | string,null | no | "png" | | | `prompt` | string | yes | | | | `response_format` | string,null | no | "url" | url, b64_json, null | | `size` | string|null | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/gpt-image-1/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/gpt-image-1/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/gpt-image-1 - Try in browser: https://aihubmix.com/playground?model=gpt-image-1 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-image-1/llms.txt --- # O4 Mini Model id on AIHubMix: `o4-mini` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > o4-mini is a remarkably smart model for its speed and cost-efficiency. This allows it to support significantly higher usage limits than o3, making it a strong high-volume, high-throughput option for everyone with questions that benefit from reasoning. - Developer: OpenAI - Context window: 200,000 tokens - Input modalities: text, image, PDF - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, File search, Structured outputs, Prompt caching - Release date: 2025-04-16 - Pricing: $1.1/M input tokens, $4.4/M output tokens, $0.275/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"o4-mini","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"o4-mini","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/o4-mini - Try in browser: https://aihubmix.com/playground?model=o4-mini - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/o4-mini.479aab11.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/o4-mini/llms.txt --- # ERNIE 5.0 Thinking Exp Model id on AIHubMix: `ernie-5.0-thinking-exp` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > ERNIE 5.0 is the next-generation natively multimodal foundation model in the ERNIE family. Built on a unified multimodal architecture, it jointly learns from text, images, audio, and video to deliver broad multimodal capabilities. ERNIE 5.0 features significantly upgraded core capabilities and shows strong performance across benchmarks, with notable gains in multimodal understanding, instruction following, creative writing, factual accuracy, and agent planning with tool use. - Developer: Baidu - Context window: 119,000 tokens - Input modalities: text - Pricing: $0.8219/M input tokens, $3.2877/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"ernie-5.0-thinking-exp","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"ernie-5.0-thinking-exp","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/ernie-5.0-thinking-exp - Try in browser: https://aihubmix.com/playground?model=ernie-5.0-thinking-exp - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/ernie-5.0-thinking-exp.9b644e2f.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/ernie-5.0-thinking-exp/llms.txt --- # Gemini 2.5 Flash Image Preview Model id on AIHubMix: `gemini-2.5-flash-image-preview` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Aihubmix supports the gemini-2.5-flash-image-preview model; you can add extra parameters modalities=["text", "image"] through the OpenAI-compatible chat interface; https://docs.aihubmix.com/en/api/Gemini-Guides#gemini-2-5-flash%3A-quick-task-support - Developer: Google - Input modalities: image, text - Pricing: per-generation — see https://aihubmix.com/model/gemini-2.5-flash-image-preview ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/edits` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-2.5-flash-image-preview","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string,null | no | | 16:9, 9:16, 1:1, 4:3, 3:4, 3:2, … | | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `image` | - | no | | | | `images` | array,null | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `negative_prompt` | string,null | no | | | | `output_format` | string,null | no | "png" | png, jpeg, webp, null | | `prompt` | string | yes | | | | `response_format` | string,null | no | "url" | url, b64_json, null | | `seed` | integer,null | no | | | | `size` | string,null | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/gemini-2.5-flash-image-preview/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/gemini-2.5-flash-image-preview/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/gemini-2.5-flash-image-preview - Try in browser: https://aihubmix.com/playground?model=gemini-2.5-flash-image-preview - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-2.5-flash-image-preview/llms.txt --- # GPT 4.1 Model id on AIHubMix: `gpt-4.1` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > The latest flagship multimodal model supports million-token context, with encoding capability (SWE-bench 54.6%) and instruction-following (Scale AI 38.3%) performance significantly surpassing GPT-4o, while reducing costs by 26%, making it suitable for complex tasks. Its automatic caching mechanism offers a 75% cost reduction on cache hits. - Developer: OpenAI - Context window: 1,047,576 tokens - Input modalities: text, image, PDF - Capabilities: Streaming, Tool calling, Web search, Code interpreter, File search, Structured outputs, Prompt caching - Release date: 2025-04-14 - Pricing: $2/M input tokens, $8/M output tokens, $0.5/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-4.1","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-4.1","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-4.1 - Try in browser: https://aihubmix.com/playground?model=gpt-4.1 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-4.1.40ab7362.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-4.1/llms.txt --- # Grok 4 Model id on AIHubMix: `grok-4` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Grok, their latest and greatest flagship model, offers unparalleled performance in natural language, math, and reasoning – the perfect jack of all trades. The current pointing model version is grok-4-0709. - Developer: Grok - Context window: 256,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Release date: 2025-07-09 - Pricing: $3.3/M input tokens, $16.5/M output tokens, $0.825/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"grok-4","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"grok-4","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/grok-4 - Try in browser: https://aihubmix.com/playground?model=grok-4 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/grok-4.c5c2a827.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/grok-4/llms.txt --- # Kimi K2 Instruct Model id on AIHubMix: `kimi-k2-instruct` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Kimi-K2 is a MoE architecture foundational model with extremely powerful coding and agent capabilities, featuring a total of 1 trillion parameters and activating 32 billion parameters. In benchmark performance tests across major categories such as general knowledge reasoning, programming, mathematics, and agents, the K2 model outperforms other mainstream open-source models. The Kimi-K2 model supports a context length of 128k tokens. It does not support visual capabilities. - Developer: Moonshot AI - Input modalities: text - Capabilities: Streaming, Tool calling - Release date: 2024-01-01 - Pricing: $0.54/M input tokens, $2.16/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"kimi-k2-instruct","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"kimi-k2-instruct","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/kimi-k2-instruct - Try in browser: https://aihubmix.com/playground?model=kimi-k2-instruct - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/kimi-k2-instruct.4f03a3af.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/kimi-k2-instruct/llms.txt --- # Kimi K2 Turbo Preview Model id on AIHubMix: `kimi-k2-turbo-preview` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > The kimi-k2-turbo-preview model is a high-speed version of kimi-k2, with the same model parameters as kimi-k2, but the output speed has been increased from 10 tokens per second to 40 tokens per second. - Developer: Moonshot AI - Context window: 262,144 tokens - Input modalities: text - Capabilities: Tool calling, Prompt caching - Release date: 2025-07-08 - Pricing: $1.2/M input tokens, $4.8/M output tokens, $0.3/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"kimi-k2-turbo-preview","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"kimi-k2-turbo-preview","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/kimi-k2-turbo-preview - Try in browser: https://aihubmix.com/playground?model=kimi-k2-turbo-preview - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/kimi-k2-turbo-preview.cd454a0d.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/kimi-k2-turbo-preview/llms.txt --- # GPT 4.1 Mini Model id on AIHubMix: `gpt-4.1-mini` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Lightweight, high-performance model with million-token context and near-flagship-level encoding and image understanding capabilities, while reducing costs by 83%. It is suitable for rapid development and small to medium-sized applications. The automatic caching mechanism provides a 75% cost reduction on cache hits. - Developer: OpenAI - Context window: 1,047,576 tokens - Input modalities: text, image, PDF - Capabilities: Streaming, Tool calling, Web search, Code interpreter, File search, Structured outputs - Release date: 2025-04-14 - Pricing: $0.4/M input tokens, $1.6/M output tokens, $0.1/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-4.1-mini","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-4.1-mini","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-4.1-mini - Try in browser: https://aihubmix.com/playground?model=gpt-4.1-mini - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-4.1-mini.c612cf96.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-4.1-mini/llms.txt --- # GPT 4.1 Nano Model id on AIHubMix: `gpt-4.1-nano` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Ultra-lightweight model with million-token context, optimized for speed and low latency, costing only $0.10 per million input tokens. It is suitable for edge computing and real-time interaction. The automatic caching mechanism offers a 75% cost reduction on cache hits. - Developer: OpenAI - Context window: 1,047,576 tokens - Input modalities: text, image, PDF - Capabilities: Streaming, Tool calling, Code interpreter, File search, Structured outputs, Prompt caching - Release date: 2025-04-14 - Pricing: $0.1/M input tokens, $0.4/M output tokens, $0.025/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-4.1-nano","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-4.1-nano","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-4.1-nano - Try in browser: https://aihubmix.com/playground?model=gpt-4.1-nano - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-4.1-nano.462eda5d.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-4.1-nano/llms.txt --- # Gemini 2.5 Pro Preview 05-06 Model id on AIHubMix: `gemini-2.5-pro-preview-05-06` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > gemini-2.5-pro latest model - Developer: Google - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio, PDF - Release date: 2025-05-07 - Pricing: $1.25/M input tokens, $10/M output tokens, $0.125/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-2.5-pro-preview-05-06","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gemini-2.5-pro-preview-05-06","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gemini-2.5-pro-preview-05-06 - Try in browser: https://aihubmix.com/playground?model=gemini-2.5-pro-preview-05-06 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gemini-2.5-pro-preview-05-06.eebb411a.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-2.5-pro-preview-05-06/llms.txt --- # Doubao Seedream 4.0 Model id on AIHubMix: `doubao-seedream-4-0` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Seedream 4.0 is a SOTA-level multimodal image creation model based on leading architecture. It breaks the creative boundaries of traditional text-to-image models by natively supporting text, single image, and multiple image inputs. Users can freely combine text and images to achieve various creative styles within the same model, such as multi-image fusion creation based on subject consistency, image editing, and set image generation, making image creation more flexible and controllable. Seedream 4.0 supports composite editing with up to 10 images in a single input. Through deep reasoning of prompt words, it automatically adapts the optimal image aspect ratio and generation quantity, enabling continuous output of up to 15 content-related images at one time. Additionally, the model significantly improves the accuracy and content diversity of Chinese generation, supports 4K ultra-high-definition output, and provides a one-stop solution from generation to editing for professional image creation. - Developer: Doubao - Input modalities: text, image - Pricing: per-generation — see https://aihubmix.com/model/doubao-seedream-4-0 ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/edits` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"doubao-seedream-4-0","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string,null | no | | 16:9, 9:16, 1:1, 4:3, 3:4, 3:2, … | | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `image` | - | no | | | | `images` | array,null | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `prompt` | string | yes | | | | `response_format` | string,null | no | "url" | url, b64_json, null | | `seed` | integer,null | no | | | | `size` | string,null | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/doubao-seedream-4-0/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/doubao-seedream-4-0/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/doubao-seedream-4-0 - Try in browser: https://aihubmix.com/playground?model=doubao-seedream-4-0 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/doubao-seedream-4-0/llms.txt --- # Qwen Image Model id on AIHubMix: `qwen-image` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Qwen-Image is a foundational image generation model in the Qwen series, achieving significant progress in complex text rendering and precise image editing. Experiments show that the model has strong general capabilities in image generation and editing, especially excelling in Chinese text rendering. - Developer: Qwen - Input modalities: text, image - Release date: 2025-08-07 - Pricing: per-generation — see https://aihubmix.com/model/qwen-image ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen-image","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `negative_prompt` | string,null | no | | | | `output_format` | string,null | no | "png" | png, jpeg, webp, null | | `prompt` | string | yes | | | | `response_format` | string,null | no | "url" | url, b64_json, null | | `seed` | integer,null | no | | | | `size` | string,null | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/qwen-image/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/qwen-image/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/qwen-image - Try in browser: https://aihubmix.com/playground?model=qwen-image - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen-image/llms.txt --- # Qwen Image Edit Model id on AIHubMix: `qwen-image-edit` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Qwen-Image-Edit is the image editing version of Qwen-Image. Based on the 20B Qwen-Image model, Qwen-Image-Edit successfully extends Qwen-Image's unique text rendering capabilities to image editing tasks, achieving precise text editing. Additionally, Qwen-Image-Edit can input the same image into Qwen2.5-VL (for visual semantic control) and the VAE encoder (for visual appearance control), enabling both semantic and appearance editing functionalities. - Developer: Qwen - Input modalities: text, image - Release date: 2025-08-19 - Pricing: per-generation — see https://aihubmix.com/model/qwen-image-edit ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/edits` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen-image-edit","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `image` | - | no | | | | `images` | array,null | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `negative_prompt` | string,null | no | | | | `output_format` | string,null | no | "png" | png, jpeg, webp, null | | `prompt` | string | yes | | | | `response_format` | string,null | no | "url" | url, b64_json, null | | `seed` | integer,null | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/qwen-image-edit/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/qwen-image-edit/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/qwen-image-edit - Try in browser: https://aihubmix.com/playground?model=qwen-image-edit - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen-image-edit/llms.txt --- # Qwen Image Max Model id on AIHubMix: `qwen-image-max` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Qwen-Image-Edit is the image editing version of Qwen-Image. Based on the 20B Qwen-Image model, Qwen-Image-Edit successfully extends Qwen-Image's unique text rendering capabilities to image editing tasks, achieving precise text editing. Additionally, Qwen-Image-Edit can input the same image into Qwen2.5-VL (for visual semantic control) and the VAE encoder (for visual appearance control), enabling both semantic and appearance editing functionalities. - Developer: Qwen - Input modalities: text, image - Pricing: per-generation — see https://aihubmix.com/model/qwen-image-max ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen-image-max","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `negative_prompt` | string,null | no | | | | `output_format` | string,null | no | "png" | png, jpeg, webp, null | | `prompt` | string | yes | | | | `response_format` | string,null | no | "url" | url, b64_json, null | | `seed` | integer,null | no | | | | `size` | string,null | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/qwen-image-max/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/qwen-image-max/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/qwen-image-max - Try in browser: https://aihubmix.com/playground?model=qwen-image-max - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen-image-max/llms.txt --- # MiMo V2 Flash (free) Model id on AIHubMix: `mimo-v2-flash-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > MiMo-V2-Flash is an open-source foundation language model developed by Xiaomi. It adopts a MoE architecture with 309B total parameters and 15B active parameters per inference, balancing performance and efficiency. The model features a hybrid attention architecture, supports a hybrid-thinking toggle, and offers a 256K context window, enabling strong capabilities in complex reasoning, code generation, and agent-based scenarios. On SWE-bench Verified and SWE-bench Multilingual, MiMo-V2-Flash ranks #1 among open-source models globally, delivering performance comparable to Claude Sonnet 4.5 while costing only about 3.5% as much. **Free tier of `mimo-v2-flash`.** The capability list and parameter schema below are those of `mimo-v2-flash` — the two ids are served by the same upstream model through the same gateway, so what the model can do is identical. What differs is quota: the free tier is rate- and volume-limited, and can return HTTP 429 where the paid id would not. Send requests with `"model":"mimo-v2-flash-free"`, not `"mimo-v2-flash"`. For the paid id's own document see https://aihubmix.com/model/mimo-v2-flash/llms.txt. - Developer: Xiaomi - Context window: 256,000 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Release date: 2025-12-16 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"mimo-v2-flash-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"mimo-v2-flash-free","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/mimo-v2-flash-free - Try in browser: https://aihubmix.com/playground?model=mimo-v2-flash-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/mimo-v2-flash.56748769.json — per-protocol parameters with types, ranges, enums and defaults, published under `mimo-v2-flash` (same upstream model). Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find `mimo-v2-flash`, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/mimo-v2-flash-free/llms.txt --- # GPT 4o Mini Tts Model id on AIHubMix: `gpt-4o-mini-tts` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > OpenAI’s latest TTS model, gpt-4o-mini-tts, uses the same API endpoint (/v1/audio/speech) as tts-1. However, OpenAI introduced a new pricing method without providing billing details via API, causing discrepancies between official pricing and aihubmix’s charges—some requests may cost more, others less. Avoid using this model if precise billing accuracy is essential. - Developer: OpenAI - Input modalities: text - Pricing: $0.6/M input tokens, $12/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-4o-mini-tts","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-4o-mini-tts","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-4o-mini-tts - Try in browser: https://aihubmix.com/playground?model=gpt-4o-mini-tts - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-4o-mini-tts.9d1105c4.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-4o-mini-tts/llms.txt --- # GPT 4o Model id on AIHubMix: `gpt-4o` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > GPT-4o (“o” stands for “omni”) is a new-generation multimodal model designed for more natural human–computer interaction. It can accept any combination of text, audio, image, and video as input, and generate multimodal outputs including text, audio, and images. With audio response latency as low as 232 milliseconds on average around 320 milliseconds, it approaches real human conversational speed. The model delivers strong performance in English text and code, significantly improved multilingual understanding, and outstanding capabilities in visual and audio perception, while offering faster API performance and substantially reduced cost for real-time and complex multimodal applications. - Developer: OpenAI - Context window: 128,000 tokens - Input modalities: text, image, PDF - Capabilities: Streaming, Tool calling, Web search, Code interpreter, File search, Structured outputs - Release date: 2024-05-13 - Pricing: $2.5/M input tokens, $10/M output tokens, $1.25/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-4o","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"gpt-4o","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-4o - Try in browser: https://aihubmix.com/playground?model=gpt-4o - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-4o.787c6431.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-4o/llms.txt --- # Command A 03 2025 Model id on AIHubMix: `command-a-03-2025` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > Command A is Cohere most performant model to date, excelling at tool use, agents, retrieval augmented generation (RAG), and multilingual use cases. Command A has a context length of 256K, only requires two GPUs to run, and has 150% higher throughput compared to Command R+ 08-2024. - Developer: Cohere - Context window: 256,000 tokens - Input modalities: text - Capabilities: Tool calling, Structured outputs - Release date: 2025-03-13 - Pricing: $2.5/M input tokens, $10/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"command-a-03-2025","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"command-a-03-2025","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/command-a-03-2025 - Try in browser: https://aihubmix.com/playground?model=command-a-03-2025 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/command-a-03-2025.f6e95d84.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/command-a-03-2025/llms.txt --- # Wan2.6 T2i Model id on AIHubMix: `wan2.6-t2i` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms - Developer: Qwen - Input modalities: image, text - Pricing: per-generation — see https://aihubmix.com/model/wan2.6-t2i ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"wan2.6-t2i","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string,null | no | | 16:9, 9:16, 1:1, 4:3, 3:4, 3:2, … | | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `output_format` | string,null | no | "png" | png, jpeg, webp, null | | `prompt` | string | yes | | | | `response_format` | string,null | no | "url" | url, b64_json, null | | `seed` | integer,null | no | | | | `size` | string,null | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/wan2.6-t2i/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/wan2.6-t2i/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/wan2.6-t2i - Try in browser: https://aihubmix.com/playground?model=wan2.6-t2i - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/wan2.6-t2i/llms.txt --- # Qwen3.8 27B Model id on AIHubMix: `qwen3.8-27b` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > qwen3.8-27b is Qwen's open-weight dense vision-language model, designed for coding, professional workflows, research, multimodal interaction, and long-running agent tasks. It supports text and image inputs with an extensive 1,000,000-token context length and flexible thinking capabilities. - Developer: Qwen - Context window: 1,000,000 tokens - Input modalities: text, image, video - Capabilities: Thinking, Tool calling, Web search, Code interpreter, Structured outputs, Prompt caching - Pricing: $0.2/M input tokens, $2.5/M output tokens, $0.05/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen3.8-27b","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"qwen3.8-27b","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/qwen3.8-27b - Try in browser: https://aihubmix.com/playground?model=qwen3.8-27b - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/qwen3.8-27b.f45d364d.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen3.8-27b/llms.txt --- # Solar Pro4 Model id on AIHubMix: `solar-pro4` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms - Developer: Upstage - Context window: 524,288 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Structured outputs - Release date: 2026-08-06 - Pricing: $0.3/M input tokens, $1.2/M output tokens, $0.06/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"solar-pro4","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"solar-pro4","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/solar-pro4 - Try in browser: https://aihubmix.com/playground?model=solar-pro4 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/solar-pro4.451947af.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/solar-pro4/llms.txt --- # Tts 1 Model id on AIHubMix: `tts-1` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms - Developer: OpenAI - Input modalities: text - Release date: 2023-11-06 - Pricing: $15/M input tokens, $15/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"tts-1","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"tts-1","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/tts-1 - Try in browser: https://aihubmix.com/playground?model=tts-1 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/tts-1.2dbf0475.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/tts-1/llms.txt --- # Tts 1 Hd Model id on AIHubMix: `tts-1-hd` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms - Developer: OpenAI - Input modalities: text - Release date: 2023-11-06 - Pricing: $30/M input tokens, $30/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"tts-1-hd","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response Without `stream`, `/v1/chat/completions` returns a standard Chat Completions object: ```json {"id":"...","object":"chat.completion","model":"tts-1-hd","choices":[{"message":{"role":"assistant","content":"..."}}],"usage":{"prompt_tokens":12,"completion_tokens":24,"total_tokens":36}} ``` With `"stream": true` the response is `text/event-stream`: read each `data:` JSON chunk until `data: [DONE]`. The Messages and Gemini endpoints return their protocols' native response shapes (Anthropic / Google). ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/tts-1-hd - Try in browser: https://aihubmix.com/playground?model=tts-1-hd - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/tts-1-hd.87f41de1.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/tts-1-hd/llms.txt --- # Qwen3.8 Flash (free) Model id on AIHubMix: `qwen3.8-flash-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > Qwen3.8 Flash is available for free on AIHubMix as part of GMI Cloud’s promotional campaign. The offer lasts for 30 days from launch or until the campaign credits are exhausted, whichever comes first. - Developer: Qwen - Context window: 1,000,000 tokens - Input modalities: text, image, video - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Structured outputs, Prompt caching - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen3.8-flash-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model qwen3.8-flash-free has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/qwen3.8-flash-free - Try in browser: https://aihubmix.com/playground?model=qwen3.8-flash-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/qwen3.8-flash-free.15eb0cc5.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen3.8-flash-free/llms.txt --- # Qwen3.8 Max (free) Model id on AIHubMix: `qwen3.8-max-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > Qwen3.8 Max is available for free on AIHubMix as part of GMI Cloud’s promotional campaign. The offer lasts for 30 days from launch or until the campaign credits are exhausted, whichever comes first. - Developer: Qwen - Context window: 1,000,000 tokens - Input modalities: text, image, video - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Structured outputs, Prompt caching - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen3.8-max-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model qwen3.8-max-free has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/qwen3.8-max-free - Try in browser: https://aihubmix.com/playground?model=qwen3.8-max-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/qwen3.8-max-free.9a1279be.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen3.8-max-free/llms.txt --- # Gemini 3.8 Flash (free) Model id on AIHubMix: `gemini-3.8-flash-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > Gemini 3.8 Flash free version: fFree model resources are limited and provided only for trial use; stability cannot be guaranteed, and you may encounter 429 errors during use. If you need to use it in a production environment and require unlimited concurrency with absolute stability, please choose the official version: gemini-3.8-flash - Developer: Google - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio, PDF - Capabilities: Thinking, Streaming, Tool calling, Web search, URL context, Code interpreter, Computer use, File search, Structured outputs, Prompt caching - Release date: 2026-08-13 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /gemini/v1beta/models/gemini-3.8-flash-free:generateContent` — Google Gemini (`x-goog-api-key: $AIHUBMIX_API_KEY`; streaming: `:streamGenerateContent?alt=sse`) - `POST /v1/messages` — Anthropic Messages (`x-api-key: $AIHUBMIX_API_KEY` + `anthropic-version: 2023-06-01`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-3.8-flash-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model gemini-3.8-flash-free has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gemini-3.8-flash-free - Try in browser: https://aihubmix.com/playground?model=gemini-3.8-flash-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gemini-3.8-flash-free.caa42f80.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-3.8-flash-free/llms.txt --- # Hy3 (free) Model id on AIHubMix: `hy3-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model was retired on 2026-07-21.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > The Hy3 official version is honed for real-world business scenarios, using a Mixture-of-Experts (MoE) architecture with 295B total parameters and 21B activated parameters. It natively supports a 256K context window and offers multiple thinking modes: no_think (ultra-fast response), think_low (quick thinking), and think_high (deep reasoning), balancing ultra-fast responses, complex reasoning, and invocation cost. Compared with the Preview version, Hy3—based on real business feedback from Tencent Yuanbao, WorkBuddy, ima, Marvis, and others—focuses on improving the Coding Agent, long-form understanding, multi-turn context continuity, search QA, and complex task execution, performing more stably in reducing hallucinations, improving task completion, and engineering usability. It is better suited to practical scenarios such as frontend tasks, cross-file code development, long-document analysis, office automation, and multi-step Agent workflows. - Developer: Hunyuan - Context window: 256,000 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2026-07-06 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"hy3-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model hy3-free has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/hy3-free - Try in browser: https://aihubmix.com/playground?model=hy3-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/hy3-free.2a6553fa.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/hy3-free/llms.txt --- # Gemini 3.7 Flash (free) Model id on AIHubMix: `gemini-3.7-flash-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > Gemini 3.7 Flash free version: fFree model resources are limited and provided only for trial use; stability cannot be guaranteed, and you may encounter 429 errors during use. If you need to use it in a production environment and require unlimited concurrency with absolute stability, please choose the official version: gemini-3.7-flash - Developer: Google - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio, PDF - Capabilities: Thinking, Streaming, Tool calling, Web search, URL context, Code interpreter, Computer use, File search, Structured outputs, Prompt caching - Release date: 2026-08-13 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /gemini/v1beta/models/gemini-3.7-flash-free:generateContent` — Google Gemini (`x-goog-api-key: $AIHUBMIX_API_KEY`; streaming: `:streamGenerateContent?alt=sse`) - `POST /v1/messages` — Anthropic Messages (`x-api-key: $AIHUBMIX_API_KEY` + `anthropic-version: 2023-06-01`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-3.7-flash-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model gemini-3.7-flash-free has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gemini-3.7-flash-free - Try in browser: https://aihubmix.com/playground?model=gemini-3.7-flash-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gemini-3.7-flash-free.b565c45b.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-3.7-flash-free/llms.txt --- # MiniMax M2.7 (free) Model id on AIHubMix: `minimax-m2.7-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model was retired on 2026-09-07.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > Developed by Minimax, MiniMax-M2.7 is a next-generation large language model designed for autonomous, real-world productivity and continuous improvement. Featuring a generous 196,608-token context window, the model integrates advanced agentic capabilities through multi-agent workflows. It is uniquely built to actively participate in its own evolution, delivering highly adaptable and intelligent performance. - Developer: Minimax - Context window: 204,800 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Prompt caching - Release date: 2026-03-18 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"minimax-m2.7-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model minimax-m2.7-free has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/minimax-m2.7-free - Try in browser: https://aihubmix.com/playground?model=minimax-m2.7-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/minimax-m2.7-free.a517aea1.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/minimax-m2.7-free/llms.txt --- # Ox Alpha Model id on AIHubMix: `ox-alpha` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > This model actually points to glm-5.3-flash; if you need to use it in production, you can directly call the model name "glm-5.3-flash". - Developer: Z.AI - Context window: 1,048,576 tokens - Input modalities: text, image, video - Capabilities: Thinking, Streaming, Tool calling, Prompt caching - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"ox-alpha","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model ox-alpha has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/ox-alpha - Try in browser: https://aihubmix.com/playground?model=ox-alpha - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/ox-alpha.ce6416a8.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/ox-alpha/llms.txt --- # Ling 3.0 Tiny (free) Model id on AIHubMix: `ling-3.0-tiny-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model was retired on 2026-10-23.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > Ling 3.0 Tiny is a mixture-of-experts model from InclusionAI, featuring 1.3B active parameters out of a total of 7.9B. It is designed for responsive agents, instruction following, and multi-turn conversations, supporting an extensive context length of 262,144 tokens. This makes it an efficient and powerful choice for conversational AI applications. - Developer: Inclusionai - Context window: 262,144 tokens - Input modalities: text - Capabilities: Tool calling - Release date: 2026-08-06 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"ling-3.0-tiny-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model ling-3.0-tiny-free has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/ling-3.0-tiny-free - Try in browser: https://aihubmix.com/playground?model=ling-3.0-tiny-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/ling-3.0-tiny-free.e37d2bff.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/ling-3.0-tiny-free/llms.txt --- # Ling 3.0 Flash (free) Model id on AIHubMix: `ling-3.0-flash-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model was retired on 2026-08-11.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > Developed by Inclusionai, ling-3.0-flash-free is a 124B-parameter Mixture-of-Experts (MoE) model with approximately 5.1B parameters activated per token. Featuring an expansive context length of 262,144 tokens, this model is built to handle extensive datasets and long-form content. It is designed with token efficiency and production-scale agentic inference as key priorities to enable seamless developer deployment. - Developer: InclusionAI - Context window: 262,144 tokens - Input modalities: text - Capabilities: Thinking, Tool calling, Structured outputs - Release date: 2026-07-23 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"ling-3.0-flash-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model ling-3.0-flash-free has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/ling-3.0-flash-free - Try in browser: https://aihubmix.com/playground?model=ling-3.0-flash-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/ling-3.0-flash-free.59ff94dd.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/ling-3.0-flash-free/llms.txt --- # Qwen3.8 Max Preview Model id on AIHubMix: `qwen3.8-max-preview` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > Qwen 3.8 Max Preview(Qwen3.8-Max-Preview) is the latest-generation foundation model in the Qwen family, packing 2.4T parameters and still evolving. Compared with the previous flagship Qwen 3.7 Max, it delivers major gains in core capabilities like Coding and Cowork (professional productivity), with world-leading performance on complex, long-horizon tasks such as full-stack development, data analysis, and Office workflows. Launch offer: Credits are consumed at just 20% of the standard rate, effectively 5× your usage. Limited time only. - Developer: Qwen - Context window: 1,000,000 tokens - Input modalities: text, image, video - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Structured outputs, Prompt caching - Release date: 2026-07-19 - Pricing: $0.338/M input tokens, $1.014/M output tokens, $0.0338/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen3.8-max-preview","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model qwen3.8-max-preview has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/qwen3.8-max-preview - Try in browser: https://aihubmix.com/playground?model=qwen3.8-max-preview - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/qwen3.8-max-preview.942b2c05.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen3.8-max-preview/llms.txt --- # Gemini 3.5 Flash Lite (free) Model id on AIHubMix: `gemini-3.5-flash-lite-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > Gemini 3.5 Flash-Lite free version: Free model resources are limited and provided only for trial use; stability cannot be guaranteed, and you may encounter 429 errors during use. If you need to use it in a production environment and require unlimited concurrency with absolute stability, please choose the official version:gemini-3.5-flash-lite. - Developer: Google - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio, PDF - Capabilities: Thinking, Tool calling, Web search, URL context, Code interpreter, File search, Structured outputs, Prompt caching - Release date: 2026-07-21 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /gemini/v1beta/models/gemini-3.5-flash-lite-free:generateContent` — Google Gemini (`x-goog-api-key: $AIHUBMIX_API_KEY`; streaming: `:streamGenerateContent?alt=sse`) - `POST /v1/messages` — Anthropic Messages (`x-api-key: $AIHUBMIX_API_KEY` + `anthropic-version: 2023-06-01`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-3.5-flash-lite-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model gemini-3.5-flash-lite-free has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gemini-3.5-flash-lite-free - Try in browser: https://aihubmix.com/playground?model=gemini-3.5-flash-lite-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gemini-3.5-flash-lite-free.2eec0ec2.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-3.5-flash-lite-free/llms.txt --- # Gemini 3.6 Flash (free) Model id on AIHubMix: `gemini-3.6-flash-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > Gemini 3.6 Flash free version: fFree model resources are limited and provided only for trial use; stability cannot be guaranteed, and you may encounter 429 errors during use. If you need to use it in a production environment and require unlimited concurrency with absolute stability, please choose the official version: gemini-3.6-flash - Developer: Google - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio, PDF - Capabilities: Thinking, Tool calling, Web search, URL context, Code interpreter, Computer use, File search, Structured outputs, Prompt caching - Release date: 2026-07-21 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /gemini/v1beta/models/gemini-3.6-flash-free:generateContent` — Google Gemini (`x-goog-api-key: $AIHUBMIX_API_KEY`; streaming: `:streamGenerateContent?alt=sse`) - `POST /v1/messages` — Anthropic Messages (`x-api-key: $AIHUBMIX_API_KEY` + `anthropic-version: 2023-06-01`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-3.6-flash-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model gemini-3.6-flash-free has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gemini-3.6-flash-free - Try in browser: https://aihubmix.com/playground?model=gemini-3.6-flash-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gemini-3.6-flash-free.d3e3e238.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-3.6-flash-free/llms.txt --- # GPT Oss 20B (free) Model id on AIHubMix: `gpt-oss-20b-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > Developed by OpenAI, gpt-oss-20b-free is an open-weight 21B parameter model released under the Apache 2.0 license. This model utilizes a Mixture-of-Experts (MoE) architecture with 3.6B active parameters per forward pass. It supports an expansive context window of up to 131,072 tokens, making it well-suited for long-context tasks. - Developer: OpenAI - Context window: 131,072 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Structured outputs - Release date: 2025-08-05 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-oss-20b-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model gpt-oss-20b-free has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-oss-20b-free - Try in browser: https://aihubmix.com/playground?model=gpt-oss-20b-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-oss-20b-free.414d41f2.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-oss-20b-free/llms.txt --- # Nemotron Nano 9B V2 (free) Model id on AIHubMix: `nemotron-nano-9b-v2-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model was retired on 2026-08-24.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > NVIDIA-Nemotron-Nano-9B-v2-free is a large language model trained from scratch by NVIDIA. Designed as a unified model, it efficiently handles both reasoning and non-reasoning tasks to respond to a wide range of user queries. With a generous context length of 128,000 tokens, it is highly capable of processing long and complex documents. - Developer: Nvidia - Context window: 131,072 tokens - Input modalities: text - Capabilities: Thinking, Tool calling - Release date: 2025-08-18 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"nemotron-nano-9b-v2-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model nemotron-nano-9b-v2-free has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/nemotron-nano-9b-v2-free - Try in browser: https://aihubmix.com/playground?model=nemotron-nano-9b-v2-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/nemotron-nano-9b-v2-free.7e799568.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/nemotron-nano-9b-v2-free/llms.txt --- # Hy3 Preview Model id on AIHubMix: `hy3-preview` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > Hunyuan Hy3 preview is designed for agent workloads, adopting a MoE architecture with 295B capacity and 21B activated parameters. It provides three modes within the same model—no_think (ultra-fast response), think_low (fast thinking), and think_high (deep reasoning)—to accommodate different latency and depth requirements from high-frequency interactions to complex engineering tasks. On code benchmarks such as SWE-bench Verified it approaches the current state of the art, and its 256K context supports cross-file code refactoring and long-document analysis. It is suitable for developers who require reliable task completion while being sensitive to inference costs. - Developer: Hunyuan - Context window: 256,000 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Structured outputs, Prompt caching - Release date: 2026-04-23 - Pricing: $0.17/M input tokens, $0.5667/M output tokens, $0.051/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"hy3-preview","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model hy3-preview has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/hy3-preview - Try in browser: https://aihubmix.com/playground?model=hy3-preview - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/hy3-preview.bd66da1b.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/hy3-preview/llms.txt --- # Nemotron Nano 12B V2 VL (free) Model id on AIHubMix: `nemotron-nano-12b-v2-vl-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model was retired on 2026-08-24.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > Developed by Nvidia, Nemotron-Nano-12B-V2-VL-Free is a 12-billion-parameter open multimodal reasoning model designed for video understanding and document intelligence. It features a powerful 128,000-token context length to handle extensive visual and textual inputs. The model utilizes a hybrid Transformer-Mamba architecture, successfully combining transformer-level accuracy with Mamba's structural advantages. - Developer: Nvidia - Context window: 131,072 tokens - Input modalities: text, image, video - Capabilities: Thinking, Streaming, Tool calling - Release date: 2025-10-28 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"nemotron-nano-12b-v2-vl-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model nemotron-nano-12b-v2-vl-free has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/nemotron-nano-12b-v2-vl-free - Try in browser: https://aihubmix.com/playground?model=nemotron-nano-12b-v2-vl-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/nemotron-nano-12b-v2-vl-free.ef7b7565.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/nemotron-nano-12b-v2-vl-free/llms.txt --- # Nemotron 3 Super 120B A12B (free) Model id on AIHubMix: `nemotron-3-super-120b-a12b-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model was retired on 2026-10-25.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > NVIDIA Nemotron 3 Super is a 120B-parameter open hybrid MoE model built on a hybrid Mamba-Transformer architecture. Activating just 12B parameters, it delivers maximum compute efficiency and accuracy for complex multi-agent applications. Additionally, it features an expansive context length of 262,144 tokens to handle large-scale inputs seamlessly. - Developer: Nvidia - Context window: 1,048,576 tokens - Input modalities: text - Capabilities: Thinking, Tool calling, Structured outputs - Release date: 2026-03-11 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"nemotron-3-super-120b-a12b-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model nemotron-3-super-120b-a12b-free has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/nemotron-3-super-120b-a12b-free - Try in browser: https://aihubmix.com/playground?model=nemotron-3-super-120b-a12b-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/nemotron-3-super-120b-a12b-free.faed3122.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/nemotron-3-super-120b-a12b-free/llms.txt --- # Laguna Xs 2.1 (free) Model id on AIHubMix: `laguna-xs-2.1-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model was retired on 2026-10-31.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > Laguna XS 2.1 is the latest coding agent model in the 33B-A3B category from Poolside, representing a step forward from the Laguna XS.2 model released in April 2026. It features an extensive context length of 262,144 tokens, allowing it to process large amounts of code and developer inputs. This model provides a robust solution for a wide range of coding agent tasks. - Developer: Poolside - Context window: 262,144 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling - Release date: 2026-07-02 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"laguna-xs-2.1-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model laguna-xs-2.1-free has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/laguna-xs-2.1-free - Try in browser: https://aihubmix.com/playground?model=laguna-xs-2.1-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/laguna-xs-2.1-free.55d513a5.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/laguna-xs-2.1-free/llms.txt --- # Laguna S 2.1 (free) Model id on AIHubMix: `laguna-s-2.1-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model was retired on 2026-10-31.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > Laguna S 2.1 is the latest coding agent model from Poolside, featuring an impressive context length of 262,144 tokens. This model is built with 118B total parameters and 8B active parameters, balancing efficiency with high performance. It delivers strong capabilities for developer tasks, scoring 70.2% on the Terminal-Bench 2.1 benchmark. - Developer: Poolside - Context window: 1,048,576 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling - Release date: 2026-07-21 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"laguna-s-2.1-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model laguna-s-2.1-free has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/laguna-s-2.1-free - Try in browser: https://aihubmix.com/playground?model=laguna-s-2.1-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/laguna-s-2.1-free.88ac2ed7.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/laguna-s-2.1-free/llms.txt --- # MiMo V2.5 Model id on AIHubMix: `mimo-v2.5` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model was retired on 2026-10-21.** Requests return 404 model_retired. Use `mimo-v2.6-flash` instead. Its guide: https://aihubmix.com/model/mimo-v2.6-flash/llms.txt > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > MiMo-V2.5 is a native, fully multimodal large model designed for agent scenarios; it can see, hear, and read, and translate understanding into action. It has over 1 trillion total parameters (42B active parameters), employs an innovative hybrid-attention architecture, and supports an ultra-long 1M context length. Built on a powerful model base, we continuously scale compute across broader agent scenarios, further expanding the agent’s action space and achieving an important generalization from coding to claw. - Developer: Xiaomi - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Release date: 2026-04-22 - Pricing: $0.155/M input tokens, $0.31/M output tokens, $0.0031/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"mimo-v2.5","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model mimo-v2.5 has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/mimo-v2.5 - Try in browser: https://aihubmix.com/playground?model=mimo-v2.5 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/mimo-v2.5.e90d651f.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/mimo-v2.5/llms.txt --- # MiMo V2.5 Pro Model id on AIHubMix: `mimo-v2.5-pro` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model was retired on 2026-10-21.** Requests return 404 model_retired. Use `mimo-v2.6-pro` instead. Its guide: https://aihubmix.com/model/mimo-v2.6-pro/llms.txt > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > MiMo-V2.5-Pro is Xiaomi's most powerful model to date. In areas such as general agent capabilities, complex software engineering, and long-horizon tasks, it can now directly compete with the world's top agent models (Claude Opus 4.6, GPT-5.4). Compared with the previous-generation MiMo-V2-Pro, it achieves an all-around leap forward. - Developer: Xiaomi - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Release date: 2026-04-22 - Pricing: $0.48/M input tokens, $0.96/M output tokens, $0.0038/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"mimo-v2.5-pro","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model mimo-v2.5-pro has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/mimo-v2.5-pro - Try in browser: https://aihubmix.com/playground?model=mimo-v2.5-pro - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/mimo-v2.5-pro.23a9e410.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/mimo-v2.5-pro/llms.txt --- # Qwen3.6 Max Preview Model id on AIHubMix: `qwen3.6-max-preview` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > The Max model Preview version, the largest and most capable model in the Qwen3.6 series, currently offers its pure-text model capabilities for trial. Compared with the previously released Qwen3-Max and Qwen3.6-Plus, this model further enhances vibe coding capabilities, executes coding agents more efficiently, significantly improves front-end programming and development capabilities, and further upgrades long-tail knowledge handling. - Developer: Qwen - Context window: 262,144 tokens - Input modalities: text - Release date: 2026-04-20 - Pricing: $1.268/M input tokens, $7.608/M output tokens, $0.1268/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen3.6-max-preview","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model qwen3.6-max-preview has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/qwen3.6-max-preview - Try in browser: https://aihubmix.com/playground?model=qwen3.6-max-preview - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/qwen3.6-max-preview.801e27bf.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen3.6-max-preview/llms.txt --- # Nemotron 3 Nano 30B A3B (free) Model id on AIHubMix: `nemotron-3-nano-30b-a3b-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model was retired on 2026-08-24.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > NVIDIA Nemotron 3 Nano 30B A3B is a highly efficient small language Mixture of Experts (MoE) model developed by Nvidia. Designed to help developers build specialized agentic AI systems, it delivers exceptional compute efficiency and accuracy. Additionally, it features an impressive context length of 256,000 tokens to support extensive data processing. - Developer: Nvidia - Context window: 1,048,576 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling - Release date: 2025-12-15 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"nemotron-3-nano-30b-a3b-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model nemotron-3-nano-30b-a3b-free has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/nemotron-3-nano-30b-a3b-free - Try in browser: https://aihubmix.com/playground?model=nemotron-3-nano-30b-a3b-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/nemotron-3-nano-30b-a3b-free.7e838ee8.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/nemotron-3-nano-30b-a3b-free/llms.txt --- # Coding Xiaomi Mimo V2.5 Model id on AIHubMix: `coding-xiaomi-mimo-v2.5` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > Only supports OpenAI-compatible formats. - Developer: Xiaomi - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Release date: 2026-04-22 - Pricing: $0.08/M input tokens, $0.16/M output tokens, $0.0016/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"coding-xiaomi-mimo-v2.5","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model coding-xiaomi-mimo-v2.5 has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/coding-xiaomi-mimo-v2.5 - Try in browser: https://aihubmix.com/playground?model=coding-xiaomi-mimo-v2.5 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/coding-xiaomi-mimo-v2.5.f8fa0393.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/coding-xiaomi-mimo-v2.5/llms.txt --- # Coding Xiaomi Mimo V2.5 Pro Model id on AIHubMix: `coding-xiaomi-mimo-v2.5-pro` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > Only supports OpenAI-compatible formats. - Developer: Xiaomi - Context window: 1,048,576 tokens - Input modalities: text - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Release date: 2026-04-22 - Pricing: $0.2/M input tokens, $0.4/M output tokens, $0.0016/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"coding-xiaomi-mimo-v2.5-pro","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model coding-xiaomi-mimo-v2.5-pro has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/coding-xiaomi-mimo-v2.5-pro - Try in browser: https://aihubmix.com/playground?model=coding-xiaomi-mimo-v2.5-pro - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/coding-xiaomi-mimo-v2.5-pro.04c7ae2f.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/coding-xiaomi-mimo-v2.5-pro/llms.txt --- # GPT 5.5 (free) Model id on AIHubMix: `gpt-5.5-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > This free model API comes from the OpenAI model deployed on Azure. To prevent abuse, the external content filter provided by Azure has been enforced, which will result in additional delays. If you want to experience the full version of the model API without filters, please use the paid version and request the model name ID without "-free". To maintain reliable service, each account is limited to 5 requests per minute, 100 requests per day, and 1 million tokens per day. - Developer: OpenAI - Context window: 1,050,000 tokens - Input modalities: text, image, PDF - Capabilities: Thinking, Streaming, Tool calling, Web search, Code interpreter, Computer use, File search, Structured outputs, Prompt caching, Server-side sessions - Release date: 2026-04-23 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-5.5-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model gpt-5.5-free has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-5.5-free - Try in browser: https://aihubmix.com/playground?model=gpt-5.5-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-5.5-free.94c9f38e.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-5.5-free/llms.txt --- # Coding Xiaomi Mimo V2 Omni Model id on AIHubMix: `coding-xiaomi-mimo-v2-omni` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > Only supports OpenAI-compatible formats. - Developer: Xiaomi - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Pricing: $0.08/M input tokens, $0.4/M output tokens, $0.016/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"coding-xiaomi-mimo-v2-omni","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model coding-xiaomi-mimo-v2-omni has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/coding-xiaomi-mimo-v2-omni - Try in browser: https://aihubmix.com/playground?model=coding-xiaomi-mimo-v2-omni - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/coding-xiaomi-mimo-v2-omni.8699a3af.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/coding-xiaomi-mimo-v2-omni/llms.txt --- # Coding Xiaomi Mimo V2 Pro Model id on AIHubMix: `coding-xiaomi-mimo-v2-pro` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > Only supports OpenAI-compatible formats. - Developer: Xiaomi - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Pricing: $0.2/M input tokens, $0.6/M output tokens, $0.04/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"coding-xiaomi-mimo-v2-pro","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model coding-xiaomi-mimo-v2-pro has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/coding-xiaomi-mimo-v2-pro - Try in browser: https://aihubmix.com/playground?model=coding-xiaomi-mimo-v2-pro - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/coding-xiaomi-mimo-v2-pro.c748beef.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/coding-xiaomi-mimo-v2-pro/llms.txt --- # GPT Image 2 (free) Model id on AIHubMix: `gpt-image-2-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > This free model API comes from the OpenAI model deployed on Azure. To prevent abuse, the external content filter provided by Azure has been enforced, which will result in additional delays. If you want to experience the full version of the model API without filters, please use the paid version and request the model name ID without "-free". - Developer: OpenAI - Input modalities: text, image - Release date: 2026-04-21 - Pricing: per-generation — see https://aihubmix.com/model/gpt-image-2-free ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/edits` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-image-2-free","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `image` | - | no | | | | `images` | array,null | no | | | | `mask` | - | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `output_format` | string,null | no | "png" | | | `prompt` | string | yes | | | | `size` | string,null | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/gpt-image-2-free/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/gpt-image-2-free/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/gpt-image-2-free - Try in browser: https://aihubmix.com/playground?model=gpt-image-2-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-image-2-free/llms.txt --- # Xiaomi Mimo V2 Omni (free) Model id on AIHubMix: `xiaomi-mimo-v2-omni-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > xiaomi-mimo-v2-omni-free is the open free version of xiaomi-mimo-v2-omni. To maintain reliable service, each account is limited to 5 requests per minute, 100 requests per day, and 1 million tokens per day. - Developer: Xiaomi - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"xiaomi-mimo-v2-omni-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model xiaomi-mimo-v2-omni-free has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/xiaomi-mimo-v2-omni-free - Try in browser: https://aihubmix.com/playground?model=xiaomi-mimo-v2-omni-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/xiaomi-mimo-v2-omni-free.6f1ea57e.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/xiaomi-mimo-v2-omni-free/llms.txt --- # Xiaomi Mimo V2 Pro (free) Model id on AIHubMix: `xiaomi-mimo-v2-pro-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > xiaomi-mimo-v2-pro-free is the open free version of xiaomi-mimo-v2-pro. To maintain reliable service, each account is limited to 5 requests per minute, 500 requests per day, and 1 million tokens per day. - Developer: Xiaomi - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"xiaomi-mimo-v2-pro-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model xiaomi-mimo-v2-pro-free has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/xiaomi-mimo-v2-pro-free - Try in browser: https://aihubmix.com/playground?model=xiaomi-mimo-v2-pro-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/xiaomi-mimo-v2-pro-free.39849fde.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/xiaomi-mimo-v2-pro-free/llms.txt --- # Xiaomi Mimo V2.5 (free) Model id on AIHubMix: `xiaomi-mimo-v2.5-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > xiaomi-mimo-v2.5-free is the open free version of xiaomi-mimo-v2.5. To maintain reliable service, each account is limited to 5 requests per minute, 100 requests per day, and 1 million tokens per day. - Developer: Xiaomi - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Release date: 2026-04-22 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"xiaomi-mimo-v2.5-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model xiaomi-mimo-v2.5-free has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/xiaomi-mimo-v2.5-free - Try in browser: https://aihubmix.com/playground?model=xiaomi-mimo-v2.5-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/xiaomi-mimo-v2.5-free.d11fd595.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/xiaomi-mimo-v2.5-free/llms.txt --- # Xiaomi Mimo V2.5 Pro (free) Model id on AIHubMix: `xiaomi-mimo-v2.5-pro-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > xiaomi-mimo-v2.5-pro-free is the open free version of xiaomi-mimo-v2.5-pro. To maintain reliable service, each account is limited to 5 requests per minute, 100 requests per day, and 1 million tokens per day. - Developer: Xiaomi - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Release date: 2026-04-22 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"xiaomi-mimo-v2.5-pro-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model xiaomi-mimo-v2.5-pro-free has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/xiaomi-mimo-v2.5-pro-free - Try in browser: https://aihubmix.com/playground?model=xiaomi-mimo-v2.5-pro-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/xiaomi-mimo-v2.5-pro-free.c3b585df.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/xiaomi-mimo-v2.5-pro-free/llms.txt --- # Gemini 3.1 Flash Image Preview (free) Model id on AIHubMix: `gemini-3.1-flash-image-preview-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > This model is the free trial version of gemini-3.1-flash-image-preview (officially marketed as Nano Banana 2). To ensure stable service operation, usage limits have been set and resources are limited and may not be available at all times. For a stable version, please choose the official release; the request name for the official release is: gemini-3.1-flash-image-preview. - Developer: Google - Input modalities: text, image - Release date: 2026-02-26 - Pricing: per-generation — see https://aihubmix.com/model/gemini-3.1-flash-image-preview-free ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/edits` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-3.1-flash-image-preview-free","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string,null | no | | 16:9, 9:16, 1:1, 4:3, 3:4, 3:2, … | | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `image` | - | no | | | | `images` | array,null | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `negative_prompt` | string,null | no | | | | `output_format` | string,null | no | "png" | png, jpeg, webp, null | | `prompt` | string | yes | | | | `response_format` | string,null | no | "url" | url, b64_json, null | | `seed` | integer,null | no | | | | `size` | string,null | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/gemini-3.1-flash-image-preview-free/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/gemini-3.1-flash-image-preview-free/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/gemini-3.1-flash-image-preview-free - Try in browser: https://aihubmix.com/playground?model=gemini-3.1-flash-image-preview-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-3.1-flash-image-preview-free/llms.txt --- # MiMo V2 Omni Model id on AIHubMix: `mimo-v2-omni` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > MiMo-V2-Omni is designed for complex real-world multimodal interaction and execution scenarios. We've built an all-modal foundation from the ground up that fuses text, vision, and speech, and use a unified architecture to deeply bind "perception" and "action." This not only breaks the traditional models' limitation of emphasizing understanding over execution, but also natively equips the model with multimodal perception, tool invocation, function execution, and GUI operation capabilities. MiMo-V2-Omni can seamlessly integrate with major agent frameworks, achieving a leap from understanding to manipulation and significantly lowering the barrier to deploying full-modal agents. - Developer: Xiaomi - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Release date: 2026-03-18 - Pricing: $0.44/M input tokens, $2.2/M output tokens, $0.088/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"mimo-v2-omni","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model mimo-v2-omni has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/mimo-v2-omni - Try in browser: https://aihubmix.com/playground?model=mimo-v2-omni - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/mimo-v2-omni.97c10db9.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/mimo-v2-omni/llms.txt --- # MiMo V2 Pro Model id on AIHubMix: `mimo-v2-pro` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > Xiaomi MiMo-V2-Pro is built for high-intensity agent work scenarios in the real world. It has over 1 trillion total parameters (42B active parameters), employs an innovative hybrid-attention architecture, and supports an ultra-long 1M-token context length. On top of a powerful model foundation, we continuously scale compute across broader agent scenarios, further expanding the intelligent action space and achieving significant generalization from Coding to Claw. - Developer: Xiaomi - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Release date: 2026-03-18 - Pricing: $1.1/M input tokens, $3.3/M output tokens, $0.22/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"mimo-v2-pro","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model mimo-v2-pro has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/mimo-v2-pro - Try in browser: https://aihubmix.com/playground?model=mimo-v2-pro - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/mimo-v2-pro.e91e5ded.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/mimo-v2-pro/llms.txt --- # Gemini 3 Flash Preview (free) Model id on AIHubMix: `gemini-3-flash-preview-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > gemini-3-flash-preview-free is the free, publicly available version of gemini-3-flash-preview, offering the same model capabilities with usage limits in place to ensure service stability. Limits include up to 5 requests per minute, a maximum of 250 requests per day, and a daily quota of 500,000 tokens. Free usage is based on shared capacity and is limited in availability. This version is intended for testing and light usage; for consistent and reliable access, please switch to the paid model. - Developer: Google - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio, PDF - Capabilities: Thinking, Tool calling, Web search, URL context, Code interpreter, Computer use, File search, Structured outputs, Prompt caching - Release date: 2025-12-17 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-3-flash-preview-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model gemini-3-flash-preview-free has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gemini-3-flash-preview-free - Try in browser: https://aihubmix.com/playground?model=gemini-3-flash-preview-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gemini-3-flash-preview-free.d823438d.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-3-flash-preview-free/llms.txt --- # GPT 4.1 (free) Model id on AIHubMix: `gpt-4.1-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > This free model API comes from the OpenAI model deployed on Azure. To prevent abuse, the external content filter provided by Azure has been enforced, which will result in additional delays. If you want to experience the full version of the model API without filters, please use the paid version and request the model name ID without "-free". - Developer: OpenAI - Context window: 1,047,576 tokens - Input modalities: text, image, PDF - Capabilities: Streaming, Tool calling, Web search, Code interpreter, File search, Structured outputs, Prompt caching - Release date: 2025-04-14 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-4.1-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model gpt-4.1-free has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-4.1-free - Try in browser: https://aihubmix.com/playground?model=gpt-4.1-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-4.1-free.23c63444.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-4.1-free/llms.txt --- # GPT 4.1 Mini (free) Model id on AIHubMix: `gpt-4.1-mini-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > This free model API comes from the OpenAI model deployed on Azure. To prevent abuse, the external content filter provided by Azure has been enforced, which will result in additional delays. If you want to experience the full version of the model API without filters, please use the paid version and request the model name ID without "-free". - Developer: OpenAI - Context window: 1,047,576 tokens - Input modalities: text, image, PDF - Capabilities: Streaming, Tool calling, Web search, Code interpreter, File search, Structured outputs - Release date: 2025-04-14 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-4.1-mini-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model gpt-4.1-mini-free has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-4.1-mini-free - Try in browser: https://aihubmix.com/playground?model=gpt-4.1-mini-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-4.1-mini-free.8d38283d.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-4.1-mini-free/llms.txt --- # GPT 4.1 Nano (free) Model id on AIHubMix: `gpt-4.1-nano-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > This free model API comes from the OpenAI model deployed on Azure. To prevent abuse, the external content filter provided by Azure has been enforced, which will result in additional delays. If you want to experience the full version of the model API without filters, please use the paid version and request the model name ID without "-free". - Developer: OpenAI - Context window: 1,047,576 tokens - Input modalities: text, image, PDF - Capabilities: Streaming, Tool calling, Code interpreter, File search, Structured outputs, Prompt caching - Release date: 2025-04-14 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-4.1-nano-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model gpt-4.1-nano-free has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-4.1-nano-free - Try in browser: https://aihubmix.com/playground?model=gpt-4.1-nano-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-4.1-nano-free.be60721d.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-4.1-nano-free/llms.txt --- # GPT 4o (free) Model id on AIHubMix: `gpt-4o-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > This free model API comes from the OpenAI model deployed on Azure. To prevent abuse, the external content filter provided by Azure has been enforced, which will result in additional delays. If you want to experience the full version of the model API without filters, please use the paid version and request the model name ID without "-free". - Developer: OpenAI - Context window: 128,000 tokens - Input modalities: text, image, PDF - Capabilities: Streaming, Tool calling, Web search, Code interpreter, File search, Structured outputs - Release date: 2024-05-13 - Pricing: free tier / see model page ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-4o-free","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model gpt-4o-free has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-4o-free - Try in browser: https://aihubmix.com/playground?model=gpt-4o-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-4o-free.3221b2b6.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-4o-free/llms.txt --- # Veo 3.1 Lite Generate Preview Model id on AIHubMix: `veo-3.1-lite-generate-preview` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. Use `gemini-omni-1.1-flash` instead. Its guide: https://aihubmix.com/model/gemini-omni-1.1-flash/llms.txt > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > Veo 3.1 is Google's state-of-the-art model for generating high-fidelity, 8-second 720p , 1080p videos featuring stunning realism and natively generated audio. - Developer: Google - Input modalities: text, image - Release date: 2026-03-31 - Pricing: per-generation — see https://aihubmix.com/model/veo-3.1-lite-generate-preview ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /ai/v1/videos/{id}` until a terminal state - `POST /v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /v1/videos/{id}` until a terminal state ## Example ```bash curl -s https://aihubmix.com/ai/v1/videos \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"veo-3.1-lite-generate-preview","prompt":""}' ``` The response contains a task id — poll `GET https://aihubmix.com/ai/v1/videos/` until `status` is terminal, then read the output URL(s). Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (video endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string|null | no | "16:9" | | | `duration` | - | no | 8 | | | `extra` | object,null | no | | | | `frame_images` | array,null | no | | | | `model` | string | yes | | | | `prompt` | string | yes | | | | `resolution` | string|null | no | | | | `seed` | integer,null | no | | | | `webhook_url` | string,null | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/veo-3.1-lite-generate-preview/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/veo-3.1-lite-generate-preview/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/veo-3.1-lite-generate-preview - Try in browser: https://aihubmix.com/playground?model=veo-3.1-lite-generate-preview - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/veo-3.1-lite-generate-preview/llms.txt --- # Doubao Seed 1.8 Model id on AIHubMix: `doubao-seed-1-8` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > Doubao's strongest multimodal Agent model Seed1.8 has powerful multimodal capabilities, supports image and text input, and can efficiently and accurately complete tasks in scenarios such as information retrieval, code generation, GUI interaction, and complex workflows, meeting increasingly diverse technical demands. - Developer: ByteDance - Context window: 256,000 tokens - Input modalities: text, image, video, document, audio - Capabilities: Thinking, Tool calling, Web search, File search, Document input, Structured outputs, Prompt caching, Server-side sessions - Release date: 2025-12-28 - Pricing: $0.1096/M input tokens, $0.274/M output tokens, $0.0219/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"doubao-seed-1-8","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model doubao-seed-1-8 has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/doubao-seed-1-8 - Try in browser: https://aihubmix.com/playground?model=doubao-seed-1-8 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/doubao-seed-1-8.af583b00.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/doubao-seed-1-8/llms.txt --- # MiMo V2 Flash Model id on AIHubMix: `mimo-v2-flash` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > MiMo-V2-Flash is a mixture of experts (MoE) language model with a total of 309 billion parameters and 15 billion activated parameters. It is designed for high-speed inference and proxy workflows, adopting a novel hybrid attention architecture and multi-token prediction (MTP), significantly reducing inference costs while achieving state-of-the-art performance. - Developer: Xiaomi - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Release date: 2025-12-16 - Pricing: $0.1918/M input tokens, $0.5754/M output tokens, $0.0384/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"mimo-v2-flash","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model mimo-v2-flash has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/mimo-v2-flash - Try in browser: https://aihubmix.com/playground?model=mimo-v2-flash - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/mimo-v2-flash.56748769.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/mimo-v2-flash/llms.txt --- # DeepSeek V3.2 Thinking Model id on AIHubMix: `deepseek-v3.2-think` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > DeepSeek-V3.2 is an efficient large language model equipped with DeepSeek Sparse Attention and reinforced reasoning performance, but its core strength lies in powerful agentic capabilities—enabled by large-scale task-synthesis that tightly integrates reasoning with real-world tool use, delivering robust, compliant, and generalizable agent behaviour. Users can toggle deeper reasoning through the reasoning_enabled switch. - Developer: DeepSeek - Context window: 163,840 tokens - Input modalities: text - Capabilities: Thinking, Tool calling, Structured outputs - Release date: 2025-12-01 - Pricing: $0.302/M input tokens, $0.453/M output tokens, $0.0302/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"deepseek-v3.2-think","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model deepseek-v3.2-think has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/deepseek-v3.2-think - Try in browser: https://aihubmix.com/playground?model=deepseek-v3.2-think - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/deepseek-v3.2-think.1093559a.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/deepseek-v3.2-think/llms.txt --- # Qwen3 VL Flash Model id on AIHubMix: `qwen3-vl-flash` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > The Qwen3 series of compact visual-understanding models achieves an effective fusion of thinking mode and non-thinking mode, outperforming the open-source Qwen3-VL-30B-A3B with faster response speeds. It comprehensively upgrades image and video understanding, supporting ultra-long contexts such as long videos and long documents, spatial awareness, and universal object recognition; it also possesses visual 2D/3D localization capabilities and is capable of handling complex real-world tasks. - Developer: Qwen - Context window: 262,144 tokens - Input modalities: text, image, video - Release date: 2025-10-09 - Pricing: $0.0206/M input tokens, $0.206/M output tokens, $0.0041/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen3-vl-flash","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model qwen3-vl-flash has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/qwen3-vl-flash - Try in browser: https://aihubmix.com/playground?model=qwen3-vl-flash - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/qwen3-vl-flash.d95e3317.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen3-vl-flash/llms.txt --- # Qwen3 VL Flash 2026 01-22 Model id on AIHubMix: `qwen3-vl-flash-2026-01-22` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > The Qwen3 series of compact visual-understanding models achieves an effective fusion of thinking mode and non-thinking mode, outperforming the open-source Qwen3-VL-30B-A3B with faster response speeds. It comprehensively upgrades image and video understanding, supporting ultra-long contexts such as long videos and long documents, spatial awareness, and universal object recognition; it also possesses visual 2D/3D localization capabilities and is capable of handling complex real-world tasks. - Developer: Qwen - Context window: 262,144 tokens - Input modalities: text, image, video - Pricing: $0.0206/M input tokens, $0.206/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen3-vl-flash-2026-01-22","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model qwen3-vl-flash-2026-01-22 has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/qwen3-vl-flash-2026-01-22 - Try in browser: https://aihubmix.com/playground?model=qwen3-vl-flash-2026-01-22 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/qwen3-vl-flash-2026-01-22.fb1efeb8.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen3-vl-flash-2026-01-22/llms.txt --- # GPT 4o Audio Preview Model id on AIHubMix: `gpt-4o-audio-preview` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > OpenAI voice input and output model, with prices consistent with the official ones. For now, only the text portion prices are displayed; voice prices can be found on the official OpenAI website. Backend billing is the same as the official. - Developer: OpenAI - Context window: 128,000 tokens - Input modalities: text, audio - Capabilities: Streaming, Tool calling, Audio output, Multimodal output - Pricing: $2.5/M input tokens, $10/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gpt-4o-audio-preview","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model gpt-4o-audio-preview has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gpt-4o-audio-preview - Try in browser: https://aihubmix.com/playground?model=gpt-4o-audio-preview - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gpt-4o-audio-preview.11a056d1.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gpt-4o-audio-preview/llms.txt --- # Gemini 2.5 Flash Model id on AIHubMix: `gemini-2.5-flash` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > Gemini 2.5 Flash is Google’s best model in terms of both performance and cost efficiency, offering a comprehensive set of capabilities. It is the first Flash model to support visible reasoning, allowing insight into the thought process behind its responses. With its strong price–performance ratio, the model is well suited for large-scale processing, low-latency, high-throughput tasks that require reasoning, as well as agent-based application scenarios. - Developer: Google - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio, PDF - Capabilities: Thinking, Tool calling, Web search, URL context, Code interpreter, File search, Structured outputs, Prompt caching - Release date: 2025-06-17 - Pricing: $0.3/M input tokens, $2.499/M output tokens, $0.03/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-2.5-flash","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model gemini-2.5-flash has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gemini-2.5-flash - Try in browser: https://aihubmix.com/playground?model=gemini-2.5-flash - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gemini-2.5-flash.a00a37d8.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-2.5-flash/llms.txt --- # Gemini 2.5 Flash Preview 09 2025 Model id on AIHubMix: `gemini-2.5-flash-preview-09-2025` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > This latest 2.5 Flash model comes with improvements in two key areas we heard consistent feedback on: Better agentic tool use: We've improved how the model uses tools, leading to better performance in more complex, agentic and multi-step applications. This model shows noticeable improvements on key agentic benchmarks, including a 5% gain on SWE-Bench Verified, compared to our last release (48.9% → 54%). More efficient: With thinking on, the model is now significantly more cost-efficient—achieving higher quality outputs while using fewer tokens, reducing latency and cost (see charts above). - Developer: Google - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio - Capabilities: Thinking, Tool calling, Web search, URL context, Code interpreter, File search, Structured outputs, Prompt caching - Release date: 2025-09-25 - Pricing: $0.3/M input tokens, $2.499/M output tokens, $0.03/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-2.5-flash-preview-09-2025","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model gemini-2.5-flash-preview-09-2025 has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gemini-2.5-flash-preview-09-2025 - Try in browser: https://aihubmix.com/playground?model=gemini-2.5-flash-preview-09-2025 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gemini-2.5-flash-preview-09-2025.d37eb5ea.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-2.5-flash-preview-09-2025/llms.txt --- # Gemini 2.5 Flash Lite Model id on AIHubMix: `gemini-2.5-flash-lite` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > Gemini 2.5 Flash-Lite is a balanced model from Google, optimized for applications that require low-latency performance. It retains the practical capabilities of the Gemini 2.5 family, including configurable reasoning based on budget, integration with tools such as grounding via Google Search and code execution, multimodal input support, and an ultra-long context window of up to 1 million tokens, delivering a strong balance between efficiency, functionality, and cost. - Developer: Google - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio, PDF - Capabilities: Thinking, Tool calling, Web search, URL context, Code interpreter, File search, Structured outputs, Prompt caching - Release date: 2025-06-17 - Pricing: $0.1/M input tokens, $0.4/M output tokens, $0.01/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-2.5-flash-lite","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model gemini-2.5-flash-lite has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gemini-2.5-flash-lite - Try in browser: https://aihubmix.com/playground?model=gemini-2.5-flash-lite - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gemini-2.5-flash-lite.57cc1165.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-2.5-flash-lite/llms.txt --- # Gemini 2.5 Flash Lite (no think) Model id on AIHubMix: `gemini-2.5-flash-lite-nothink` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > Gemini 2.5 Flash-Lite is a balanced model from Google, optimized for applications that require low-latency performance. It retains the practical capabilities of the Gemini 2.5 family, including configurable reasoning based on budget, integration with tools such as grounding via Google Search and code execution, multimodal input support, and an ultra-long context window of up to 1 million tokens, delivering a strong balance between efficiency, functionality, and cost. - Developer: Google - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio, PDF - Capabilities: Thinking, Tool calling, Web search, URL context, Code interpreter, File search, Structured outputs, Prompt caching - Pricing: $0.1/M input tokens, $0.4/M output tokens, $0.01/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-2.5-flash-lite-nothink","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model gemini-2.5-flash-lite-nothink has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gemini-2.5-flash-lite-nothink - Try in browser: https://aihubmix.com/playground?model=gemini-2.5-flash-lite-nothink - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gemini-2.5-flash-lite-nothink.825a4795.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-2.5-flash-lite-nothink/llms.txt --- # Gemini 2.5 Flash Lite Preview 09 2025 Model id on AIHubMix: `gemini-2.5-flash-lite-preview-09-2025` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > gemini-2.5-flash-lite latest preview version - Developer: Google - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio, PDF - Capabilities: Thinking, Tool calling, Web search, URL context, Code interpreter, File search, Structured outputs, Prompt caching - Release date: 2025-09-25 - Pricing: $0.1/M input tokens, $0.4/M output tokens, $0.01/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-2.5-flash-lite-preview-09-2025","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model gemini-2.5-flash-lite-preview-09-2025 has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gemini-2.5-flash-lite-preview-09-2025 - Try in browser: https://aihubmix.com/playground?model=gemini-2.5-flash-lite-preview-09-2025 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gemini-2.5-flash-lite-preview-09-2025.20989a9e.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-2.5-flash-lite-preview-09-2025/llms.txt --- # Gemini 2.5 Flash Lite Preview 09 2025 (no think) Model id on AIHubMix: `gemini-2.5-flash-lite-preview-09-2025-nothink` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > gemini-2.5-flash-lite latest preview version - Developer: Google - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio, PDF - Capabilities: Thinking, Tool calling, Web search, URL context, Code interpreter, File search, Structured outputs, Prompt caching - Pricing: $0.1/M input tokens, $0.4/M output tokens, $0.01/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-2.5-flash-lite-preview-09-2025-nothink","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model gemini-2.5-flash-lite-preview-09-2025-nothink has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gemini-2.5-flash-lite-preview-09-2025-nothink - Try in browser: https://aihubmix.com/playground?model=gemini-2.5-flash-lite-preview-09-2025-nothink - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gemini-2.5-flash-lite-preview-09-2025-nothink.5bd760b3.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-2.5-flash-lite-preview-09-2025-nothink/llms.txt --- # Gemini 2.5 Flash (no think) Model id on AIHubMix: `gemini-2.5-flash-nothink` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > Gemini-2.5-flash defaults to thinking enabled; to disable thinking, request the name gemini-2.5-flash-nothink, which only supports OpenAI-compatible format calls and does not support Gemini SDK; for the native Gemini SDK, please set the parameter budget=0 directly. - Developer: Google - Context window: 1,048,576 tokens - Input modalities: text, image, video, audio, PDF - Capabilities: Thinking, Tool calling, Web search, URL context, Code interpreter, File search, Structured outputs, Prompt caching - Release date: 2025-06-24 - Pricing: $0.3/M input tokens, $2.499/M output tokens, $0.03/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"gemini-2.5-flash-nothink","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model gemini-2.5-flash-nothink has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/gemini-2.5-flash-nothink - Try in browser: https://aihubmix.com/playground?model=gemini-2.5-flash-nothink - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/gemini-2.5-flash-nothink.fb8532b6.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/gemini-2.5-flash-nothink/llms.txt --- # DeepSeek-OCR Model id on AIHubMix: `DeepSeek-OCR` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > DeepSeek-OCR is a vision-language model launched by DeepSeek AI, focusing on optical character recognition (OCR) and “contextual optical compression.” The model is designed to explore the limits of compressing contextual information from images, efficiently processing documents and converting them into structured text formats such as Markdown. The model requires an image as input. - Developer: DeepSeek - Context window: 8,192 tokens - Input modalities: text, image - Release date: 2025-10-20 - Pricing: $0.02/M input tokens, $0.02/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"DeepSeek-OCR","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model DeepSeek-OCR has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/DeepSeek-OCR - Try in browser: https://aihubmix.com/playground?model=DeepSeek-OCR - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/DeepSeek-OCR.0bcfcd99.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/DeepSeek-OCR/llms.txt --- # grok-4-fast-non-reasoning Model id on AIHubMix: `grok-4-fast-non-reasoning` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > Grok-4-fast is a cost-effective inference model developed by xAI that delivers cutting-edge performance with excellent token efficiency. The model features a 2 million token context window, advanced Web and X search capabilities, and a unified architecture supporting both "inference" and "non-inference" modes. Compared to Grok 4, it reduces thinking tokens by an average of 40% and lowers the price by 98% while achieving the same performance. - Developer: Grok - Context window: 1,000,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Release date: 2025-07-09 - Pricing: $0.2/M input tokens, $0.5/M output tokens, $0.05/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"grok-4-fast-non-reasoning","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model grok-4-fast-non-reasoning has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/grok-4-fast-non-reasoning - Try in browser: https://aihubmix.com/playground?model=grok-4-fast-non-reasoning - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/grok-4-fast-non-reasoning.72de5889.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/grok-4-fast-non-reasoning/llms.txt --- # Grok 4 Fast (reasoning) Model id on AIHubMix: `grok-4-fast-reasoning` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > Grok-4-fast is a cost-effective inference model developed by xAI that delivers cutting-edge performance with excellent token efficiency. The model features a 2 million token context window, advanced Web and X search capabilities, and a unified architecture supporting both "inference" and "non-inference" modes. Compared to Grok 4, it reduces thinking tokens by an average of 40% and lowers the price by 98% while achieving the same performance. - Developer: Grok - Context window: 1,000,000 tokens - Input modalities: text, image - Capabilities: Thinking, Streaming, Tool calling, Web search, Structured outputs, Prompt caching - Release date: 2025-09-01 - Pricing: $0.2/M input tokens, $0.5/M output tokens, $0.05/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"grok-4-fast-reasoning","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model grok-4-fast-reasoning has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/grok-4-fast-reasoning - Try in browser: https://aihubmix.com/playground?model=grok-4-fast-reasoning - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/grok-4-fast-reasoning.f22030e4.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/grok-4-fast-reasoning/llms.txt --- # Kimi K2 0711 Model id on AIHubMix: `kimi-k2-0711` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > Kimi-K2 is a MoE architecture foundational model with extremely powerful coding and agent capabilities, featuring a total of 1 trillion parameters and activating 32 billion parameters. In benchmark performance tests across major categories such as general knowledge reasoning, programming, mathematics, and agents, the K2 model outperforms other mainstream open-source models. The Kimi-K2 model supports a context length of 128k tokens. It does not support visual capabilities. - Developer: Moonshot AI - Context window: 131,072 tokens - Input modalities: text - Capabilities: Streaming, Tool calling - Release date: 2025-01-01 - Pricing: $0.54/M input tokens, $2.16/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"kimi-k2-0711","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model kimi-k2-0711 has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/kimi-k2-0711 - Try in browser: https://aihubmix.com/playground?model=kimi-k2-0711 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/kimi-k2-0711.eee2e9f9.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/kimi-k2-0711/llms.txt --- # Qwen3 VL 235B A22B Instruct Model id on AIHubMix: `qwen3-vl-235b-a22b-instruct` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > The Qwen3 series open-source models include hybrid models, thinking models, and non-thinking models, with both reasoning capabilities and general abilities reaching industry SOTA levels at the same scale. - Developer: Qwen - Context window: 131,072 tokens - Input modalities: text, image, video - Release date: 2025-09-23 - Pricing: $0.274/M input tokens, $1.096/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen3-vl-235b-a22b-instruct","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model qwen3-vl-235b-a22b-instruct has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/qwen3-vl-235b-a22b-instruct - Try in browser: https://aihubmix.com/playground?model=qwen3-vl-235b-a22b-instruct - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/qwen3-vl-235b-a22b-instruct.be86467c.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen3-vl-235b-a22b-instruct/llms.txt --- # Qwen3 VL 235B A22B Thinking Model id on AIHubMix: `qwen3-vl-235b-a22b-thinking` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > The Qwen3 series open-source models include hybrid models, thinking models, and non-thinking models, with both reasoning capabilities and general abilities reaching industry SOTA levels at the same scale. - Developer: Qwen - Context window: 131,072 tokens - Input modalities: text, image, video - Release date: 2025-09-23 - Pricing: $0.274/M input tokens, $2.74/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen3-vl-235b-a22b-thinking","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model qwen3-vl-235b-a22b-thinking has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/qwen3-vl-235b-a22b-thinking - Try in browser: https://aihubmix.com/playground?model=qwen3-vl-235b-a22b-thinking - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/qwen3-vl-235b-a22b-thinking.f7dc3f7f.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen3-vl-235b-a22b-thinking/llms.txt --- # Qwen3 VL 30B A3B Instruct Model id on AIHubMix: `qwen3-vl-30b-a3b-instruct` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > The Qwen3-VL series’ second-largest MoE model Instruct version offers fast response speed and supports ultra-long contexts such as long videos and long documents; it features comprehensive upgrades in image/video understanding, spatial perception, and universal recognition abilities; it also provides visual 2DD/3D localization capabilities, making it capable of handling complex real-world tasks. - Developer: Qwen - Context window: 131,072 tokens - Input modalities: text, image, video - Release date: 2025-10-05 - Pricing: $0.1028/M input tokens, $0.4112/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen3-vl-30b-a3b-instruct","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model qwen3-vl-30b-a3b-instruct has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/qwen3-vl-30b-a3b-instruct - Try in browser: https://aihubmix.com/playground?model=qwen3-vl-30b-a3b-instruct - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/qwen3-vl-30b-a3b-instruct.d98ad590.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen3-vl-30b-a3b-instruct/llms.txt --- # Qwen3 VL 30B A3B Thinking Model id on AIHubMix: `qwen3-vl-30b-a3b-thinking` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > The Qwen3-VL series’ second-largest MoE model Thinking version offers fast response speed, stronger multimodal understanding and reasoning, visual agent capabilities, and ultra-long context support for long videos and long documents; it features comprehensive upgrades in image/video understanding, spatial perception, and universal recognition abilities, making it capable of handling complex real-world tasks. - Developer: Qwen - Context window: 131,072 tokens - Input modalities: text, image, video - Release date: 2025-10-11 - Pricing: $0.1028/M input tokens, $1.028/M output tokens ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen3-vl-30b-a3b-thinking","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model qwen3-vl-30b-a3b-thinking has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/qwen3-vl-30b-a3b-thinking - Try in browser: https://aihubmix.com/playground?model=qwen3-vl-30b-a3b-thinking - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/qwen3-vl-30b-a3b-thinking.48d0a581.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen3-vl-30b-a3b-thinking/llms.txt --- # Veo 3.1 Fast Generate Preview Model id on AIHubMix: `veo-3.1-fast-generate-preview` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. Use `gemini-omni-1.1-flash` instead. Its guide: https://aihubmix.com/model/gemini-omni-1.1-flash/llms.txt > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > Veo 3.1 is Google's state-of-the-art model for generating high-fidelity, 8-second 720p , 1080p or 4k videos featuring stunning realism and natively generated audio. - Developer: Google - Input modalities: text, image, video - Release date: 2025-10-15 - Pricing: per-generation — see https://aihubmix.com/model/veo-3.1-fast-generate-preview ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /ai/v1/videos/{id}` until a terminal state - `POST /v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /v1/videos/{id}` until a terminal state ## Example ```bash curl -s https://aihubmix.com/ai/v1/videos \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"veo-3.1-fast-generate-preview","prompt":""}' ``` The response contains a task id — poll `GET https://aihubmix.com/ai/v1/videos/` until `status` is terminal, then read the output URL(s). Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (video endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string|null | no | "16:9" | | | `duration` | - | no | 8 | | | `extra` | object,null | no | | | | `frame_images` | array,null | no | | | | `input_references` | array,null | no | | | | `model` | string | yes | | | | `prompt` | string | yes | | | | `resolution` | string|null | no | | | | `seed` | integer,null | no | | | | `webhook_url` | string,null | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/veo-3.1-fast-generate-preview/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/veo-3.1-fast-generate-preview/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/veo-3.1-fast-generate-preview - Try in browser: https://aihubmix.com/playground?model=veo-3.1-fast-generate-preview - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/veo-3.1-fast-generate-preview/llms.txt --- # Veo 3.1 Generate Preview Model id on AIHubMix: `veo-3.1-generate-preview` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. Use `gemini-omni-1.1-flash` instead. Its guide: https://aihubmix.com/model/gemini-omni-1.1-flash/llms.txt > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > Veo 3.1 is Google's state-of-the-art model for generating high-fidelity, 8-second 720p , 1080p or 4k videos featuring stunning realism and natively generated audio. - Developer: Google - Input modalities: text, image, video - Release date: 2025-10-15 - Pricing: per-generation — see https://aihubmix.com/model/veo-3.1-generate-preview ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /ai/v1/videos/{id}` until a terminal state - `POST /v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /v1/videos/{id}` until a terminal state ## Example ```bash curl -s https://aihubmix.com/ai/v1/videos \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"veo-3.1-generate-preview","prompt":""}' ``` The response contains a task id — poll `GET https://aihubmix.com/ai/v1/videos/` until `status` is terminal, then read the output URL(s). Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (video endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string|null | no | "16:9" | | | `duration` | - | no | 8 | | | `extra` | object,null | no | | | | `frame_images` | array,null | no | | | | `input_references` | array,null | no | | | | `model` | string | yes | | | | `prompt` | string | yes | | | | `resolution` | string|null | no | | | | `seed` | integer,null | no | | | | `webhook_url` | string,null | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/veo-3.1-generate-preview/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/veo-3.1-generate-preview/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/veo-3.1-generate-preview - Try in browser: https://aihubmix.com/playground?model=veo-3.1-generate-preview - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/veo-3.1-generate-preview/llms.txt --- # Qwen3 Max Model id on AIHubMix: `qwen3-max` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > The Tongyi Qianwen 3 series Max model has undergone special upgrades in intelligent agent programming and tool invocation compared to the preview version. The officially released model this time reaches SOTA level in the field and is adapted to more complex intelligent agent scenarios. - Developer: Qwen - Context window: 262,144 tokens - Input modalities: text - Release date: 2025-09-23 - Pricing: $0.4508/M input tokens, $1.8032/M output tokens, $0.0902/M cached input ## Endpoints (base URL: https://aihubmix.com) - `POST /v1/chat/completions` — OpenAI Chat Completions (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/v1/chat/completions \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"qwen3-max","messages":[{"role":"user","content":"Hello"}]}' ``` ## Response This id is retired, so the call above no longer succeeds. It returns HTTP 404: ```json {"error":{"code":"model_retired","message":"The model qwen3-max has been retired and is no longer available. (tid: …)","type":"Aihubmix_api_error"}} ``` The request shape is kept above so it can be diffed against a successor's — point `"model"` at a live id to get a real completion. Discover live ids in the catalog (no auth): https://aihubmix.com/api/v1/models?type=llm&sort_by=order ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Model page: https://aihubmix.com/model/qwen3-max - Try in browser: https://aihubmix.com/playground?model=qwen3-max - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Full parameter schema (machine-readable, authoritative): https://aihubmix.com/model-data/models/qwen3-max.86254429.json — per-protocol parameters with types, ranges, enums and defaults. Refreshed together with this page; if it ever 404s, re-resolve via `https://aihubmix.com/model-data/index.json` (find this id, fetch its `path`) - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/qwen3-max/llms.txt --- # V3 Model id on AIHubMix: `V3` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model was retired on 2026-10-14.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > Fast and high-quality — top image quality in just 11 seconds per piece, with almost no extra time for batch generation. Flexible ratios — supports ultra-wide and tall formats like 3:1, 2:1, offering diverse perspectives. Unique strengths — outstanding design capabilities in the V3 and V2 series, with powerful text rendering (Chinese support coming soon). Precise local editing — fine-tuned mask control for area redrawing (edit) and easy background replacement (replace-background). - Developer: Ideogram - Input modalities: text, image - Pricing: per-generation — see https://aihubmix.com/model/V3 ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/edits` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"V3","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string,null | no | | 16:9, 9:16, 1:1, 4:3, 3:4, 3:2, … | | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `image` | - | no | | | | `images` | array,null | no | | | | `mask` | - | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `negative_prompt` | string,null | no | | | | `output_format` | string,null | no | "png" | png, jpeg, webp, null | | `prompt` | string | yes | | | | `response_format` | string,null | no | "url" | url, b64_json, null | | `seed` | integer,null | no | | | | `size` | string,null | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/V3/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/V3/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/V3 - Try in browser: https://aihubmix.com/playground?model=V3 - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/V3/llms.txt --- # Stable Diffusion 3.5 Large Model id on AIHubMix: `Stable-Diffusion-3-5-Large` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model was retired on 2026-10-14.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > Stable Diffusion 3.5 Large, developed by Stability AI, is a text-to-image generation model that supports high-quality image creation with excellent prompt responsiveness and customization, suitable for professional applications. - Developer: Stable diffusion - Input modalities: text, image - Pricing: per-generation — see https://aihubmix.com/model/Stable-Diffusion-3-5-Large ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/edits` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) - `POST /v1/images/generations` — image generation (`Authorization: Bearer $AIHUBMIX_API_KEY`) ## Example ```bash curl -s https://aihubmix.com/ai/v1/images/generations \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"Stable-Diffusion-3-5-Large","prompt":""}' ``` The response carries the output URL(s) directly. Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (image endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string,null | no | "1:1" | 16:9, 9:16, 1:1, 3:2, 2:3, 21:9, … | | `async` | boolean | no | false | | | `extra` | object,null | no | | | | `image` | - | no | | | | `images` | array,null | no | | | | `model` | string | yes | | | | `n` | integer,null | no | 1 | | | `negative_prompt` | string,null | no | | | | `output_format` | string,null | no | "png" | png, jpeg, webp, null | | `prompt` | string | yes | | | | `response_format` | string,null | no | "url" | url, b64_json, null | | `seed` | integer,null | no | | | | `webhook_events_filter` | array | no | | | | `webhook_url` | string | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/Stable-Diffusion-3-5-Large/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/Stable-Diffusion-3-5-Large/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/Stable-Diffusion-3-5-Large - Try in browser: https://aihubmix.com/playground?model=Stable-Diffusion-3-5-Large - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/Stable-Diffusion-3-5-Large/llms.txt --- # Wan3.0 Video (free) Model id on AIHubMix: `wan3.0-video-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > The free trial version of wan3.0-video is provided by gmicloud. - Developer: Qwen - Input modalities: text, image, audio, video, PDF - Pricing: per-generation — see https://aihubmix.com/model/wan3.0-video-free ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /ai/v1/videos/{id}` until a terminal state ## Example ```bash curl -s https://aihubmix.com/ai/v1/videos \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"wan3.0-video-free"}' ``` The response contains a task id — poll `GET https://aihubmix.com/ai/v1/videos/` until `status` is terminal, then read the output URL(s). Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (video endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string|string|null | no | "adaptive" | | | `duration` | integer|null | no | 5 | 2, 3, 4, 5, 6, 7, … | | `extra` | object,null | no | | | | `frame_images` | array | no | | | | `input_references` | array | no | | | | `model` | string | yes | | | | `prompt` | string | no | | | | `resolution` | string|null | no | "1080p" | | | `seed` | integer,null | no | | | | `webhook_url` | string,null | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/wan3.0-video-free/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/wan3.0-video-free/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/wan3.0-video-free - Try in browser: https://aihubmix.com/playground?model=wan3.0-video-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/wan3.0-video-free/llms.txt --- # Wan3.0 Video Prime (free) Model id on AIHubMix: `wan3.0-video-prime-free` Create an API key: https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms > **This model has been retired.** Requests return 404 model_retired. > > The specs and pricing below stay readable for migration and cost comparison — this id is no longer a live integration target. Responses for this id also carry RFC 8594 headers: `Sunset` (the retirement instant), `Deprecation`, and `Link: …; rel="successor-version"` when a successor is published — read those to detect this automatically. All retirements: https://aihubmix.com/models/retirements > The free trial version of wan3.0-video-prime is provided by gmicloud. - Developer: Qwen - Input modalities: text, image, audio, video, PDF - Pricing: per-generation — see https://aihubmix.com/model/wan3.0-video-prime-free ## Endpoints (base URL: https://aihubmix.com) - `POST /ai/v1/videos` — video generation (`Authorization: Bearer $AIHUBMIX_API_KEY`); async — poll `GET /ai/v1/videos/{id}` until a terminal state ## Example ```bash curl -s https://aihubmix.com/ai/v1/videos \ -H "Authorization: Bearer $AIHUBMIX_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"wan3.0-video-prime-free"}' ``` The response contains a task id — poll `GET https://aihubmix.com/ai/v1/videos/` until `status` is terminal, then read the output URL(s). Full parameter list: see the schema link below. Generation costs real money — check the price on the model page first. ## Parameters (video endpoint) | name | type | required | default | enum | |---|---|---|---|---| | `aspect_ratio` | string|string|null | no | "adaptive" | | | `duration` | integer|null | no | 5 | 2, 3, 4, 5, 6, 7, … | | `extra` | object,null | no | | | | `frame_images` | array | no | | | | `input_references` | array | no | | | | `model` | string | yes | | | | `prompt` | string | no | | | | `resolution` | string|null | no | "1080p" | | | `seed` | integer,null | no | | | | `webhook_url` | string,null | no | | | Always-current parameter schema (authoritative, no auth): ```bash curl -s https://aihubmix.com/call/schema/models/wan3.0-video-prime-free/endpoints ``` Output URLs are short-lived (roughly 30 minutes) and are downloaded with the same `Authorization: Bearer` header — fetch them promptly and store the content yourself. ## Errors Error responses carry a `tid` (trace id) — include it when contacting support. Reference: https://docs.aihubmix.com/en/FAQs/HTTP-Codes.md - 400 — parameter error; most are passed through from the upstream provider (media: `prompt_missing`, `size_not_supported`, `n_not_within_range`, …) - 401 — missing `Authorization` header, or the key is invalid/expired - 403 — `insufficient_user_quota` (top up at https://console.aihubmix.com/?utm_source=llms-agent&utm_medium=model-llms), account suspended, or this key is not allowed to use this model - 429 — rate limited; back off and retry - 503 — no channel can serve the request (check the model id and your access), or the upstream provider is throttling; retry later ## More - Machine-readable endpoint schema (no auth): https://aihubmix.com/call/schema/models/wan3.0-video-prime-free/endpoints - Async task guide: https://docs.aihubmix.com/en/api/async-tasks.md - Model page: https://aihubmix.com/model/wan3.0-video-prime-free - Try in browser: https://aihubmix.com/playground?model=wan3.0-video-prime-free - Compare with another model (human-facing, side-by-side specs and pricing): https://aihubmix.com/compare — pick this model and a peer there; published pairs are listed in https://aihubmix.com/sitemap-compare.xml, unpublished pairs 404 so do not compose the path by hand - Generate runnable code programmatically: npm `@aihubmix/codegen` — the generator behind the Playground's "Get Code" (4 protocols × 7 languages, media endpoints included); the body it builds is the exact wire body the Playground sends, so generated snippets and real requests cannot diverge. `@aihubmix/model-schema` (npm) translates the parameter schema above into codegen input - Site index for agents: https://aihubmix.com/llms.txt · Onboarding: https://aihubmix.com/agents.md --- Canonical version of this document: https://aihubmix.com/model/wan3.0-video-prime-free/llms.txt