AIHubMixAIHubMix

Llama Models

52 modelsGeneral models from $0.20/M inputUp to 1.05M context

Usage

Last 30 days · 2026-07-13 to 2026-08-11

Tokens

5.1B

Requests

374K

Models in use

22 of 52

Tokens per day

0368M736M07-1307-2007-2708-0308-102026-07-13 — 49,360,350 tokens2026-07-14 — 62,682,545 tokens2026-07-15 — 61,580,855 tokens2026-07-16 — 137,427,975 tokens2026-07-17 — 76,345,260 tokens2026-07-18 — 13,772,570 tokens2026-07-19 — 705,655,505 tokens2026-07-20 — 135,069,170 tokens2026-07-21 — 106,561,880 tokens2026-07-22 — 22,687,225 tokens2026-07-23 — 538,646,785 tokens2026-07-24 — 660,609,890 tokens2026-07-25 — 83,448,230 tokens2026-07-26 — 242,035,100 tokens2026-07-27 — 460,952,540 tokens2026-07-28 — 116,410,250 tokens2026-07-29 — 2,297,060 tokens2026-07-30 — 5,155 tokens2026-07-31 — 123,854,515 tokens2026-08-01 — 2,264,410 tokens2026-08-02 — 12,233,855 tokens2026-08-03 — 20,380 tokens2026-08-04 — 16,735 tokens2026-08-05 — 72,780 tokens2026-08-06 — 57,706,600 tokens2026-08-07 — 3,574,285 tokens2026-08-08 — 197,236,570 tokens2026-08-09 — 736,064,095 tokens2026-08-10 — 524,070,090 tokens2026-08-11 — 1,530,430 tokens

Which models that traffic went to

  1. llama-3.3-70b34.2%1.8B
  2. llama-4-maverick29.7%1.5B
  3. llama-4-scout24.2%1.2B
  4. llama3.1-8b5.7%294M
  5. llama-3.1-70b4.7%241M
  6. deepinfra-llama-3.1-8b-instant1.1%57.7M
  7. deepinfra-llama-3.3-70b-instant-turbo0.4%19.5M
  8. groq-llama-3.3-70b-versatile<0.1%1.5M
  9. 5 more models<0.1%137K

Share of 5.1B tokens. 9 models with traffic report no token counts and cannot be ranked here, including deepseek-r1-distill-qianfan-llama-8b and meta-llama-3-70b — they are in the request view.

The two views disagree on purpose: a model can take a large share of the calls and a small share of the tokens — many short requests — or the reverse. Which one matters depends on whether your cost is driven by call volume or by prompt length. Measured on AIHubMix over the last 30 days, counting the 52 model IDs listed on this page; traffic routed through upstream-specific IDs that are not in the public catalog is not included.

All 52 Llama Models

Open in model list
Llama models on AIHubMix with input and output modalities, context length, maximum output, price per million tokens including cache read and cache write rates, and measured throughput and latency.
Modalities
llama-4-maverickTakes text, vision, returns text.1.05M32K$0.20$0.20/M98 tok/s0.23 s
llama-4-scoutTakes text, vision, returns text.131K—$0.20$0.20/M2637 tok/s0.29 s
llama-3.3-70b66K8K$0.60$0.60/M3250 tok/s0.25 s
llama3-groq-8b-8192-tool-use-preview——$0.0002$0.0002/M——
llama3-groq-70b-8192-tool-use-preview——$0.0009$0.0009/M——
meta-llama/llama-3.1-405b-instruct:free——$0.02$0.02/M——
meta-llama/llama-3.1-70b-instruct:free——$0.02$0.02/M——
meta-llama/llama-3.1-8b-instruct:free——$0.02$0.02/M——
meta-llama/llama-3.2-11b-vision-instruct:free——$0.02$0.02/M——
meta-llama/llama-3.2-3b-instruct:free——$0.02$0.02/M——
deepinfra-llama-3.1-8b-instant——$0.03$0.05/M——
groq-llama-3.1-8b-instant——$0.06$0.09/M——
llama3-8b-8192——$0.06$0.12/M——
deepinfra-llama-4-scout-17b-16e-instruct——$0.09$0.33/M——
llama2-7b-2048——$0.10$0.10/M——
deepinfra-llama-3.3-70b-instant-turbo——$0.11$0.35/M——
groq-llama-4-scout-17b-16e-instruct——$0.12$0.37/M——
deepseek-r1-distill-qianfan-llama-8b——$0.14$0.55/M——
llama-3.2-11b-vision-preview——$0.20$0.20/M——
llama-3.2-1b-preview——$0.20$0.20/M——
llama-3.2-3b-preview——$0.20$0.20/M——
groq-llama-4-maverick-17b-128e-instruct——$0.22$0.66/M——
qianfan-llama-vl-8b——$0.27$0.69/M——
aihubmix-Llama-3-1-8B-Instruct——$0.30$0.60/M——
llama-3.1-8b-instant——$0.30$0.60/M——
llama3-8b-8192(33)——$0.30$0.30/M——
llama3.1-8b——$0.30$0.60/M107 tok/s0.17 s
meta/llama3-8B-chat——$0.30$0.30/M——
deepinfra-llama-4-maverick-17b-128e-instruct——$0.33$1.32/M——
aihubmix-Llama-3-2-11B-Vision——$0.40$0.40/M——
Gryphe/MythoMax-L2-13b——$0.40$0.40/M——
llama-3.1-70b——$0.44$0.44/M12 tok/s0.30 s
llama2-70b-4096Takes , returns text.——$0.50$0.50/M——
llama2-70b-40960Takes , returns text.——$0.50$0.50/M——
meta-llama/Llama-3.2-90B-Vision-Instruct——$0.50$0.50/M——
meta-llama-3-8b——$0.55$0.55/M——
aihubmix-Llama-3-1-70B-Instruct——$0.60$0.78/M——
cerebras-llama-3.3-70b——$0.60$0.60/M——
llama-3.1-70b-versatile——$0.60$0.60/M——
groq-llama-3.3-70b-versatile——$0.65$0.87/M——
aihubmix-Llama-3-70B-Instruct——$0.70$0.70/M——
llama3-70b-8192——$0.70$0.94/M——
qianfan-chinese-llama-2-13b——$0.82$0.82/M——
WizardLM/WizardCoder-Python-34B-V1.0——$0.90$0.90/M——
aihubmix-Llama-3-2-90B-Vision——$2.40$2.40/M——
llama-3.2-90b-vision-preview——$2.40$2.40/M——
llama3-70b-8192(33)——$2.65$2.65/M——
llama-3.1-405b-instruct——$4.00$4.00/M——
llama-3.1-405b-reasoning——$4.00$4.00/M——
meta-llama-3-70b——$4.79$4.79/M——
aihubmix-Llama-3-1-405B-Instruct——$5.00$15.00/M——
meta/llama-3.1-405b-instruct——$5.00$5.00/M——

Prices are USD per million tokens; cache read and cache write are the rates for prompt-cache hits and for writing a prompt into the cache. Throughput and latency are measured on AIHubMix — the same figures the model detail page shows — not vendor claims. A dash means the catalog does not publish that field for that model, which is not the same as the model not supporting it.

Common Comparisons

llama-4-maverick vs llama-4-scoutllama-4-maverick vs llama2-70b-4096llama-4-maverick vs llama2-70b-40960llama-4-maverick vs qwen3.8-max-previewkimi-k3 vs llama-4-maverick

Llama on AIHubMix

Which Llama model should I start with?

llama-4-maverick at $0.20/M input — the cheapest entry here that declares tool calling, and it carries a 1.05M context. Move up to aihubmix-Llama-3-1-405B-Instruct when answer quality matters more than cost.

Why are there several entries for the same model?

Because each row is a route you can call, not a model release. Some IDs name an upstream (azure-, alicloud-, cc-), some are the open-weight repository form (meta-llama/…), and some differ only in capitalisation, kept so older integrations keep working.

The catalog does not carry a field saying which of those a given row is, so this page does not sort them into buckets it would have to invent. Every row shows that route’s own price, context and speed — compare those directly, and open a model to see the upstreams that serve it.

Do I need a separate Llama account?

No. One AIHubMix key covers every model on this page, and switching between them is a change to the model string — billing, rate limits, and logs stay in one place.

Other Providers

Qwen
140 modelsOpenAI
132 modelsGoogle
81 modelsZ.AI
64 modelsByteDance
46 models

Start calling Llama in one line

One key, one endpoint, 844 models across 35 providers.

Get API keyRead the docs
AIHubMix

© 2023 - 2026 AIHubMix, LLC

Product

Company

Developer

Connect