Xiaomi MiMo V2.6 Pro
MiMo V2.6 Pro
The strongest large open-weights model on the independent index, under MIT, for under a dollar out.
Updated
The read
MiMo V2.6 Pro replaced V2.5 Pro in September, and Xiaomi retires the old version on 21 October. It is a sparse model of about a trillion parameters with 42 billion active, it reads text, images, video and audio, and it leads the large open-weights models on the Artificial Analysis index. The weights are MIT-licensed and the hosted API costs 87 cents per million output tokens, a small fraction of any Western model at this level.
Where it shines
- Highest Artificial Analysis score of any large open-weights model
- MIT licence on a trillion-parameter model
- Reads text, images, video and audio natively
Worth knowing first
- Self-hosting means serving a trillion-parameter model
- Hosted in China, with the data policy questions that brings
Index score
89.4
Provisional. This score is read from the product's public capability, not from the one-prompt rebuild the fully reviewed entries went through, so treat it as a placing rather than a verdict.
- Rank
- 06 of 73
- Builds
- LLMs
- Category
- Open weights
- Output
- MIT-licensed weights, plus a hosted API
- Pricing
- Free tier
- Best for
- Agentic work at the top of the open field, on weights you can modify
Scores are our own editorial judgement, weighted craft 55, speed 10, control 15 and value 20.
Also worth a look
Full index- 05GLM-5.3 FlashGLM-5.3 FlashProvisionalA fast, MIT-licensed open model that human voters rank beside the big flagships.Open weightsFree tier89.6
- 07DeepSeek V4.1 FlashDeepSeek V4.1 FlashProvisionalFrontier-adjacent answers for thirty cents per million input tokens, with the weights under MIT.Open weightsFree tier88.3
- 10GLM-5.3GLM-5.3ProvisionalThe same base as GLM-5.2, re-trained into one of the strongest open models available.Open weightsFree tier87.9
