Xiaomi model product
MiMo-V2-Flash is a powerful, efficient, and ultra-fast foundation language model that excels in reasoning, coding, and agentic scenarios. It is a Mixture-of-Experts model with 309B total parameters and 15B active parameters, featuring a hybrid attention architecture with sliding-window and full attention (5:1 ratio, 128-token window). Delivers 150 tokens/sec inference with 256k context window.
Updated Aug 10, 2026. Default version: MiMo-V2-Flash
Structured fields from the published default version.
This profile uses the latest version under this unique model that has a calculated LLMBoard score. Arena and price are excluded.
Published benchmark records for the scored version MiMo-V2-Flash.
| Arena-Hard v2 | 0.9 | 1 | 16 | 100.0% | C | |
| Tau-bench | 0.8 | 3 | 6 | 60.0% | C | |
| MRCR | 0.5 | 6 | 7 | 16.7% | C | |
| LongBench v2 | 0.6 | 9 | 17 | 50.0% | C | |
| SWE-bench Multilingual | 0.7 | 15 | 34 | 57.6% | C | |
| LiveCodeBench v6 | 0.8 | 19 | 53 | 65.4% | C | |
| Terminal-Bench | 0.3 | 19 | 25 | 25.0% | C | |
| HMMT 2025 | 0.8 | 22 | 33 | 34.4% | C | |
| MMLU-Pro | 0.8 | 22 | 129 | 83.6% | C | |
| AIME 2025 | 0.9 | 24 | 114 | 79.7% | C | |
| BrowseComp | 0.6 | 37 | 58 | 36.8% | C | |
| SWE-Bench Verified | 0.7 | 43 | 104 | 59.2% | C | |
| Terminal-Bench 2.0 | 0.4 | 46 | 49 | 6.3% | C | |
| Humanity's Last Exam | 0.2 | 54 | 92 | 41.8% | C | |
| GPQA | 0.8 | 64 | 233 | 72.8% | C |
Preference and agent-evaluation signals from published Arena datasets.
The default version has no published Arena rows, or its source alias has not been resolved.
Official vendor API PAYG pricing is summarized first. The table then lists individual provider offerings without treating their minimum as the official price.
| Qiniu | mimo-v2-flash | global | $0.10 | $0.30 | 256K | |
| Xiaomi | mimo-v2-flash | global | $0.14 | $0.28 | 262.1K |
Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.
All published versions linked to this unique model. The score columns identify the version used by the current overall ranking.
| MiMo-V2-Flash | 55.9 | 309B | 262.1K | 65.5K | Yes | MIT |
A concise description based on the published model registry.
MiMo-V2-Flash is a powerful, efficient, and ultra-fast foundation language model that excels in reasoning, coding, and agentic scenarios. It is a Mixture-of-Experts model with 309B total parameters and 15B active parameters, featuring a hybrid attention architecture with sliding-window and full attention (5:1 ratio, 128-token window). Delivers 150 tokens/sec inference with 256k context window.
Use the benchmark, Arena and pricing sections above as separate evidence. A missing field means the current data snapshot does not support that claim.
Data snapshot: 2026-08-07. Editorial model content is not available in the backend.
Open a comparison with the three ranked models immediately above and below this model.
Recommendations prioritize the same model type and family, then the closest published LLMBoard score.
Common questions about MiMo V2 Flash.
MiMo V2 Flash's default version was released on Dec 16, 2025.
MiMo V2 Flash's official API price is $0.14 per million input tokens and $0.28 per million output tokens via Xiaomi. The lowest tracked third-party offer starts at $0.10 input and $0.30 output via Qiniu.
MiMo V2 Flash is published under Xiaomi in the model registry.
The default version has a 262.1K token context window.
Yes. The default version is marked as open weight under MIT.
2 published provider offerings are linked to the default version.
Nearby ranked alternatives include GPT-OSS-120B, GPT-5.4-mini, MiniMax M1 80K.