Xiaomi
Xiaomi, the Beijing consumer electronics company, builds the MiMo family of models, and MiMo-V2.5 had already become one of the most-used open models on OpenRouter before the V2.6 series shipped on September 21, 2026. The release is two models under the MIT license: MiMo-V2.6-Pro, a 1.02 trillion parameter mixture-of-experts with 42 billion active, and MiMo-V2.6-Flash at 309 billion with 15 billion active, both natively omnimodal across text, image, video, and audio with 1 million token context. API prices held at V2.5 levels, $0.435 and $0.87 per million tokens for Pro and $0.14 and $0.28 for Flash, with a Pro-UltraSpeed tier at ten times the Pro rate. Independent evaluators rank the pair near the top of open weights: Artificial Analysis scores Pro at 46, its highest open-weights result, and the Vals Index puts both models just under 60 percent, ahead of DeepSeek V4.1 Flash. Xiaomi published the technical report, training environments, and reinforcement learning code alongside the weights, and says the Pro RL run cost about $2.62 million over under six days. The models are available in AI Studio, MiMo Code, MiMo Desktop, the MiMo Open Platform API, and OpenRouter.
Founded
2010 (Xiaomi)
Headquarters
Beijing, China
CEO
Lei Jun
Models
2 active
Key Products
Strengths
- ✓MIT-licensed open weights at trillion-parameter scale
- ✓Native text, image, video, and audio input in one model
- ✓Top open-weights score on the Artificial Analysis index
- ✓Sub-dollar output pricing on both models
- ✓Training code and environments published with the weights
Xiaomi Models
| Model | Input / 1M | Output / 1M | Context | Capabilities |
|---|---|---|---|---|
| MiMo-V2.6-Pro | 0.43 | 0.87 | 1.0M | text, vision, audio, video, tool-use, code, reasoning |
| MiMo-V2.6-Flash | 0.14 | 0.28 | 1.0M | text, vision, audio, video, tool-use, code, reasoning |
Prices per 1M tokens in USD. See the full pricing guide for detailed analysis.
Benchmark Scores
| Model | SWE-bench | MMLU-Pro | HumanEval | GPQA Diamond | MATH | OSWorld 2.0 | BrowseComp | FrontierCode v1.1 | Terminal-Bench 4.0 | Humanity's Last Exam (tools) |
|---|---|---|---|---|---|---|---|---|---|---|
| MiMo-V2.6-Pro | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | 24.8 | N/A |
| MiMo-V2.6-Flash | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | 21.2 | N/A |