MiMo-V2.6-Pro
Xiaomi MiMoReleased Sep 21, 2026Open weights · MITmimo-v2.6-pro
- Reasoning
- yes
- Context
- 1M tokens
- Max output
- 128k tokens
- Input
- text, image, video, audio
- API features
- tools, structured outputs, JSON mode, seed
- Xiaomi MiMo's API
- yes
Read Oct 23, 06:00 UTC from the hosts' published offers and OpenRouter, spec sheet from Xiaomi MiMo
Speed by provider
Output speed of MiMo-V2.6-Pro at each host, fastest first
- Xiaomi89 tok/s3.83 s to first token
- DeepInfra37 tok/s1.46 s to first token
- GMICloud37 tok/s2.92 s to first token
- Xiaomi34 tok/s2.71 s to first token
- StreamLake23 tok/s2.37 s to first token
Speed: OpenRouter, median of the last 30 min, read Oct 23, 06:00 UTC · a snapshot of real traffic before the daily read, not a daily average · bars from zero.
Open weights under MIT: you can run MiMo-V2.6-Pro on your own hardware instead of paying per token.
Memory ≈ parameters × bytes per parameter × 1.2 · cloud price: on-demand rate, running 24/7 · builds: GGUF, AWQ, MLX, GPTQ, Ollama · break-even calculator
| Provider | Input | Output | Cache ↑ | tok/s | First token | Uptime 24h | Checked |
|---|---|---|---|---|---|---|---|
| DeepInfrafp8 · 1M ctx−10% | $0.39 | $0.78 | $0.0032 | 37 | 1.46 s | 100% | Oct 23 |
| Xiaomifp8 · 1M ctx · 128K out | $0.43 | $0.87 | $0.0036 | 34 | 2.71 s | 98.2% | Oct 23 |
| Novita AIprecision not stated · 1M ctx · 128K out | $0.43 | $0.87 | $0.0036 | — | — | — | Oct 23 |
| GMICloudbf16 (fp8) · 1.05M ctx | $0.43 | $0.87 | $0.004 | 37 | 2.92 s | 98.8% | Oct 23 |
| StreamLakeprecision not read: host terms · 1M ctx · 125K out | $0.52 | $1.04 | $0.0043 | 23 | 2.37 s | 28.6% | Oct 23 |
| Xiaomifp8 · 1M ctx · 128K out | $4.35 | $8.70 | $0.036 | 89 | 3.83 s | 100% | Oct 23 |
6 of 6 providers · prices per million tokens, read Oct 23, 06:00 UTC · speed: OpenRouter, median of the last 30 min, read Oct 23, 06:00 UTC · uptime: OpenRouter, last 24 h · −N%: a discount off the host's list price · under each host: weights precision, context and max output as served, the maker's value in brackets when the host's differs
| Provider | Weights | Context | Max output | Tools | Structured outputs | Reasoning |
|---|---|---|---|---|---|---|
| DeepInfra | fp8 | 1M | not stated by the host | yes | yes | yes |
| GMICloud | bf16 | 1.05M | not read: host terms | yes | yes | yes |
| Xiaomi | fp8 | 1M | 128K | yes | yes | yes |
| Novita AI | not stated by the host | 1M | 128K | — | — | — |
| StreamLake | not read: host terms | 1M | 125K | yes | no | yes |
| Xiaomi | fp8 | 1M | 128K | yes | yes | yes |
Weights, context and max output: read at each host's own catalogue or docs where its terms allow, otherwise as the host declares to OpenRouter; hover a cell for its source and date · tools, structured outputs, reasoning: as declared to OpenRouter · not verified by us
MiMo-V2.6-Pro spec sheet, field by field, with sources
MiMo-V2.6-Pro spec sheetmaker-declared
What Xiaomi MiMo declares about the model itself — read from its own weights repository and documentation, not from any host. What each endpoint actually serves is in the pricing above: the two can differ, and both are true.
- Context window
- 1,048,576 tokens
- Parameters
- 1,024,216,603,392≈ 1.02T
- Active experts
- 8 of 384 active per token
- Weight precision
- fp8
- License
- MIT — open weights, conditionalScope: the licence on the published weights, and nothing else. It is not the maker's acceptable-use policy, which is a separate document, and it is not the contract you sign with whichever host you call — hosts set their own terms, which this index does not read.
- Knowledge cutoff
- not published by the maker
- Modalities
- text · image · video · audio
- Max output (official)
- not published by the maker
- Release date
- September 21, 2026
- Maker lifecycle status
- undefined
Also declared in the weights
- Architecture class
- MiMoV2ForCausalLM
- Layers
- 70
- Vocabulary
- 152,576 tokens
- Weights first published
- September 21, 2026
- Weights last updated
- September 22, 2026
- Languages declared
- English · Chinese
XiaomiMiMo/MiMo-V2.6-Pro-RL · config.json @ 73875d0 · read 2026-10-08 · 8 of 10 fields published by the maker
| Date | Provider | Input | Output | Change |
|---|---|---|---|---|
| Oct 22 | StreamLake | $0.43 → $0.52 | $0.87 → $1.04 | +20% |
| Oct 20 | DeepInfra | $0.43 → $0.39 | $0.87 → $0.78 | −10% |
| Oct 4 | GMICloud | $0.41 → $0.43 | $0.83 → $0.87 | +5% |
| Oct 1 | GMICloud | $0.43 → $0.41 | $0.87 → $0.83 | −5% |
| Sep 29 | DeepInfra | $0.43 → $0.43 | $0.87 → $0.87 | 0% |
Every published change of input or output price, newest first · change on input + output · full series: pricing-history.csv.
| Provider | Chatbot 1,000 in / 500 out | RAG / long context 8,000 in / 500 out | Batch summarize 4,000 in / 1,000 out |
|---|---|---|---|
| DeepInfra | $0.78 | $3.49 | $2.33 |
| GMICloud | $0.87 | $3.92 | $2.61 |
| Xiaomi | $0.87 | $3.92 | $2.61 |
| Novita AI | $0.87 | $3.92 | $2.61 |
| StreamLake | $1.04 | $4.70 | $3.13 |
| Xiaomi | $8.70 | $39.15 | $26.10 |
What each host bills for 1,000 requests of a representative job
Your own workload
Other Xiaomi MiMo models · all its models, hosts and history ›
- MiMo-V2.6-Flash9 providers · from $0.28
- MiMo-V2.5-Pro7 providers · from $0.91
- MiMo-V2.55 providers · from $0.40