Microsoft Models

17 modelsGeneral models from $0.12/M inputUp to 256K context

Usage

Last 30 days · 2026-10-22 to 2026-11-20

Tokens

638M

Requests

13.3K

Models in use

14 of 17

Tokens per day, stacked by model

0256M512M10-2210-2911-0511-1211-192026-10-22 — 92,425 tokens mai-image-2.6-flash: 41,395 mai-image-2.6: 25,615 mai-image-2.5: 7,255 mai-image-2.5-flash: 7,255 mai-image-2.5-pro: 7,255 mai-thinking-1: 3,6502026-10-23 — 12,405 tokens mai-thinking-1: 6,585 mai-image-2.6-flash: 5,8202026-10-24 — 254,315 tokens mai-image-2.6-flash: 203,705 mai-image-2.6: 45,165 mai-thinking-1: 5,4452026-10-25 — 62,745 tokens mai-image-2.6: 29,200 mai-image-2.6-flash: 27,865 mai-thinking-1: 5,6802026-10-26 — 157,380 tokens mai-image-2.6: 75,440 mai-image-2.6-flash: 74,005 mai-thinking-1: 7,9352026-10-27 — 506,955 tokens mai-thinking-1: 306,065 mai-image-2.6-flash: 189,310 mai-image-2.6: 11,5802026-10-28 — 177,405 tokens mai-image-2.6: 89,810 mai-image-2.6-flash: 82,415 mai-thinking-1: 5,1802026-10-29 — 98,470 tokens mai-image-2.6: 93,580 mai-thinking-1: 4,8902026-10-30 — 165,730 tokens mai-image-2.6: 87,640 mai-thinking-1: 78,0902026-10-31 — 2,088,235 tokens mai-thinking-1: 1,917,225 mai-image-2.6: 171,0102026-11-01 — 2,731,980 tokens mai-thinking-1: 2,536,165 mai-image-2.6: 189,810 mai-image-2.6-flash: 6,0052026-11-02 — 779,325 tokens mai-image-2.6: 773,185 mai-thinking-1: 6,1402026-11-03 — 78,454,595 tokens mai-thinking-1: 78,333,395 mai-image-2.6: 85,305 mai-image-2.6-flash: 35,8952026-11-04 — 512,025,545 tokens mai-thinking-1: 511,509,610 mai-image-2.6: 470,285 mai-image-2.6-flash: 45,6502026-11-05 — 13,171,330 tokens mai-thinking-1: 12,193,820 mai-image-2.6: 977,5102026-11-06 — 282,475 tokens mai-image-2.6: 265,275 mai-image-2.6-flash: 11,050 mai-thinking-1: 6,1502026-11-07 — 376,375 tokens mai-image-2.6: 348,630 mai-image-2.6-flash: 22,365 mai-thinking-1: 5,3802026-11-08 — 420,965 tokens mai-image-2.6: 400,325 mai-image-2.6-flash: 15,565 mai-thinking-1: 5,0752026-11-09 — 212,970 tokens mai-image-2.6: 172,075 mai-thinking-1: 23,795 mai-image-2.6-flash: 17,1002026-11-10 — 4,122,035 tokens mai-thinking-1: 4,110,035 mai-image-2.6-flash: 12,0002026-11-11 — 17,599,320 tokens mai-thinking-1: 17,599,3202026-11-12 — 1,493,205 tokens mai-thinking-1: 1,480,055 mai-image-2.6-flash: 13,1502026-11-13 — 4,805 tokens mai-thinking-1: 4,8052026-11-14 — 234,325 tokens mai-thinking-1: 234,3252026-11-15 — 208,335 tokens mai-thinking-1: 208,3352026-11-16 — 34,265 tokens mai-image-2.6-flash: 28,690 mai-thinking-1: 5,5752026-11-17 — 1,130,940 tokens mai-thinking-1: 1,130,9402026-11-18 — 149,120 tokens mai-thinking-1: 89,780 mai-image-2.6-flash: 54,105 mai-image-2.6: 5,2352026-11-19 — 107,375 tokens mai-thinking-1: 96,600 mai-image-2.6-flash: 10,7752026-11-20 — 836,820 tokens mai-thinking-1: 836,820
  • mai-thinking-1
  • mai-image-2.6
  • mai-image-2.6-flash
  • mai-image-2.5
  • mai-image-2.5-flash
  • mai-image-2.5-pro

Which models that traffic went to

  1. Mai Thinking 199.2%633M
  2. Mai Image 2.60.7%4.3M
  3. Mai Image 2.6 Flash0.1%897K
  4. Mai Image 2.5<0.1%7.3K
  5. Mai Image 2.5 Flash<0.1%7.3K
  6. Mai Image 2.5 Pro<0.1%7.3K

Share of 638M tokens. 8 models with traffic report no token counts and cannot be ranked here, including aihub-Phi-4 and aihub-Phi-4-multimodal-instruct — 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 17 model IDs listed on this page; traffic routed through upstream-specific IDs that are not in the public catalog is not included.

All 17 Microsoft Models

Open in model list
Microsoft 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
mai-thinking-1Takes text, returns text.256K64K$2$8/M78 tok/s10.08 s
aihub-Phi-4-mini-instructTakes text, returns text.128K4K$0.12$0.48/M——
aihub-Phi-4-multimodal-instructTakes text, vision, audio, returns text.128K4K$0.12$0.48/M——
AiHubmix-Phi-4-mini-reasoningTakes text, returns text.128K4K$0.12$0.12/M——
mai-image-2.6-flashTakes text, vision, returns vision.32K4K$1.75$1.75/M——
mai-image-2.6Takes text, vision, returns vision.32K4K$5$5/M——
aihub-Phi-4Takes text, returns text.16K—$0.12$0.48/M46 tok/s0.41 s
ahm-Phi-3-5-MoE-instructTakes , returns text.——$0.4$1.6/M——
ahm-Phi-3-5-vision-instructTakes text, vision, returns text.——$0.4$1.6/M——
ahm-Phi-3-5-mini-instructTakes , returns text.——$1$3/M——
ahm-Phi-3-medium-4kTakes , returns text.——$1$3/M——
ahm-Phi-3-small-128kTakes , returns text.——$1$3/M——
mai-image-2.5-flashTakes text, vision, returns vision.——$1.75$1.75/M——
mai-image-2.7-flashTakes text, vision, returns vision.——$1.75$1.75/M——
mai-image-2.5Takes text, vision, returns vision.——$5$5/M——
mai-image-2.5-proTakes text, vision, returns vision.——$5$5/M——
ahm-Phi-3-medium-128kTakes , returns text.——$6$18/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.

Microsoft on AIHubMix

Which Microsoft model should I start with?

AiHubmix-Phi-4-mini-reasoning at $0.12/M input — the cheapest entry here that declares a token price, and it carries a 128K context. Move up to ahm-Phi-3-medium-128k when answer quality matters more than cost, or to mai-thinking-1 for long-form reasoning.

Which of these models reason before answering?

1 of the 17 models here declare a reasoning phase — they work through the problem before producing an answer, which helps on multi-step problems at the cost of extra output tokens. Use the Reasoning filter above the table to see them. The catalog does not record anything further about how they differ, so this page does not sort them into families.

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-), 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 Microsoft 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.

Start calling Microsoft in one line

One key, one endpoint, 984 models across 49 model authors.