MiMo-V2.6-Flash
Xiaomi MiMoReleased Sep 21, 2026Open weights · MITmimo-v2.6-flash
- Reasoning
- yes
- Context
- 1M tokens
- Max output
- 384k tokens
- Input
- text, image, video, audio
- API features
- tools, structured outputs, JSON mode, seed, logprobs
- Xiaomi MiMo's API
- yes
Read Oct 23, 06:00 UTC from the hosts' published offers and OpenRouter, spec sheet from Xiaomi MiMo
The cheapest host, Relace, serves mxfp4 weights (maker: fp8). Cheapest at the maker's precision and context: GMICloud, $0.36 blended. Weights and context per host
Speed by provider
Output speed of MiMo-V2.6-Flash at each host, fastest first
- Cohere174 tok/s0.52 s to first token
- Relace130 tok/s0.76 s to first token
- Io Net43 tok/s1.80 s to first token
- GMICloud28 tok/s4.32 s to first token
- Xiaomi26 tok/s5.14 s to first token
- DeepInfra15 tok/s3.62 s to first token
- Venice15 tok/s6.16 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-Flash 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 |
|---|---|---|---|---|---|---|---|
| GMICloudbf16 (fp8) · 1.05M ctx−15% | $0.12 | $0.24 | $0.0026 | 28 | 4.32 s | 94.4% | Oct 23 |
| DeepInfrafp8 · 1M ctx−5% | $0.13 | $0.27 | $0.0027 | 15 | 3.62 s | 97.5% | Oct 23 |
| Xiaomifp8 · 1M ctx · 128K out | $0.14 | $0.28 | $0.0028 | 26 | 5.14 s | 100% | Oct 23 |
| Novita AIprecision not stated · 1M ctx · 128K out | $0.14 | $0.28 | $0.0028 | — | — | — | Oct 23 |
| Io Netfp8 · 1.05M ctx · 128K outreprices often | $0.10 | $0.28 | $0.003 | 43 | 1.80 s | 99.1% | Oct 23 |
| Cohereprecision not read: host terms · 1M ctx · 375K out | $0.14 | $0.28 | $0.003 | 174 | 0.52 s | 99.7% | Oct 23 |
| Venicefp8 · 1M ctx · 128K out | $0.17 | $0.35 | $0.0037 | 15 | 6.16 s | 97.8% | Oct 23 |
| Relacemxfp4 (fp8) · 1M ctx · 128K out | $0.14 | $0.14 | $0.014 | 130 | 0.76 s | 99.9% | Oct 23 |
| InferenceNet1.05M ctx | $0.12 | $0.28 | $0.11 | — | — | 99.4% | Oct 4 |
9 of 9 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 |
|---|---|---|---|---|---|---|
| Relace | mxfp4 | 1M | 128K | yes | no | yes |
| GMICloud | bf16 | 1.05M | not read: host terms | yes | yes | yes |
| Io Net | fp8 | 1.05M | 128K | yes | yes | yes |
| DeepInfra | fp8 | 1M | not stated by the host | yes | yes | yes |
| InferenceNet | not checked yet | 1.05M | not checked yet | — | — | — |
| Xiaomi | fp8 | 1M | 128K | yes | yes | yes |
| Novita AI | not stated by the host | 1M | 128K | — | — | — |
| Cohere | not read: host terms | 1M | 375K | yes | yes | yes |
| Venice | 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-Flash spec sheet, field by field, with sources
MiMo-V2.6-Flash 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
- 310,756,322,688≈ 310.8B
- Active experts
- 8 of 256 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
- 48
- Vocabulary
- 152,576 tokens
- Weights first published
- September 21, 2026
- Weights last updated
- September 22, 2026
- Languages declared
- English · Chinese
XiaomiMiMo/MiMo-V2.6-Flash-RL · config.json @ 5711b26 · read 2026-10-08 · 8 of 10 fields published by the maker
| Date | Provider | Input | Output | Change |
|---|---|---|---|---|
| Oct 23 | GMICloud | $0.14 → $0.12 | $0.28 → $0.24 | −15% |
| Oct 23 | Io Net | $0.15 → $0.10 | $0.29 → $0.28 | −14% |
| Oct 20 | Relace | $0.04 → $0.14 | $1.28 → $0.14 | −79% |
| Oct 15 | DeepInfra | $0.14 → $0.13 | $0.28 → $0.27 | −5% |
| Oct 15 | Io Net | $0.19 → $0.15 | $0.29 → $0.29 | −8% |
| Oct 12 | Io Net | $0.26 → $0.19 | $0.41 → $0.29 | −28% |
| Oct 10 | Io Net | $0.29 → $0.26 | $0.45 → $0.41 | −9% |
| Oct 9 | Io Net | $0.18 → $0.29 | $0.35 → $0.45 | +40% |
| Oct 8 | Io Net | $0.16 → $0.18 | $0.30 → $0.35 | +15% |
| Oct 7 | Io Net | $0.14 → $0.16 | $0.28 → $0.30 | +8% |
| Oct 6 | Io Net | $0.099 → $0.14 | $0.25 → $0.28 | +21% |
| Oct 5 | Io Net | $0.13 → $0.099 | $0.25 → $0.25 | −7% |
| Oct 4 | GMICloud | $0.13 → $0.14 | $0.27 → $0.28 | +5% |
| Oct 4 | Io Net | $0.14 → $0.13 | $0.27 → $0.25 | −7% |
| Sep 30 | DeepInfra | $0.12 → $0.14 | $0.24 → $0.28 | +18% |
| Sep 29 | DeepInfra | $0.14 → $0.12 | $0.28 → $0.24 | −15% |
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 |
|---|---|---|---|
| Relace | $0.21 | $1.19 | $0.70 |
| GMICloud | $0.24 | $1.07 | $0.71 |
| Io Net | $0.24 | $0.94 | $0.68 |
| DeepInfra | $0.27 | $1.20 | $0.80 |
| InferenceNet | $0.26 | $1.10 | $0.76 |
| Xiaomi | $0.28 | $1.26 | $0.84 |
| Novita AI | $0.28 | $1.26 | $0.84 |
| Cohere | $0.28 | $1.26 | $0.84 |
| Venice | $0.35 | $1.57 | $1.05 |
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-V2.5-Pro7 providers · from $0.91
- MiMo-V2.6-Pro6 providers · from $1.17
- MiMo-V2.55 providers · from $0.40