MiMo-V2-Flash
Xiaomi ·
Verified
2 of 2 published values traced to a source2 primary sources, 2 archived1 not researched yet
Why this status
Newly added. Sources are not archived and no value has been traced to one yet — run `npm run enrich`. Capabilities evidenced from the primary sources: coding: “mimo-v2-flash-0204 mimo-v2-flash-0112 mimo-v2-flash SWE-Bench Verified Non-Thinking 73.” | agentic: “Post-Training Technical Highlights MiMo-V2-Flash leverages a post-training pipeline designed to maximize reasoning and agentic capabilities through innovative distillation and rein”.
What these states mean · How records are judged · Has this record changed?
Xiaomi released MiMo-V2-Flash with a hybrid architecture using a 1:5 ratio of global attention to sliding-window attention (window size 128), native 32K context and training extended to 256K (262,144 in the published config). Xiaomi's changelog records two later updates, on 2026-01-12 and 2026-02-04, strengthening general and coding performance in Thinking mode.
Details
| Released | December 16, 2025stated in mimo.mi.com, huggingface.coMiMo-V2-Flash was released on December 16, 2025.
|
|---|---|
| Company | Xiaomi |
| Family | MiMo |
| Lineage | before MiMo-V2.6-Flash |
| Era | Reasoning models |
| Type | Language model · LLM |
| Modalities | Not recorded |
| Capabilities | Coding, Agentic |
| Weights | Open weights |
| Licence | MIT |
| Context window | 262K tokensstated in mimo.mi.com, huggingface.coMiMo-V2-Flash supports a context window of 262,144 tokens.
|
| Parameters | Not researched |
Release timeline
A model has more than one date. The timeline above positions it at its announcement; these are every recorded lifecycle event, each with the source that evidences it.
- Announced mimo.mi.com, huggingface.co
- Major update mimo.mi.com
- Major update mimo.mi.com
Current information, elsewhere
This record is a historical one: every figure above is traced to a dated primary source. For what is true today — live pricing, current rankings, availability — these are better maintained than anything here could be.
- Current price and providers OpenRouter ↗
- Current benchmark scores Artificial Analysis ↗
- Weights and model card Hugging Face ↗
Sources
Record status: verified · confidence 90/100 · 2 of 2 primary
Newly added. Sources are not archived and no value has been traced to one yet — run `npm run enrich`. Capabilities evidenced from the primary sources: coding: “mimo-v2-flash-0204 mimo-v2-flash-0112 mimo-v2-flash SWE-Bench Verified Non-Thinking 73.” | agentic: “Post-Training Technical Highlights MiMo-V2-Flash leverages a post-training pipeline designed to maximize reasoning and agentic capabilities through innovative distillation and rein”.
- mimo.mi.com Official documentation archived
- huggingface.co Model card archived
Primary means published by the organisation that made the model. Verified records require at least one.
MiMo family
- MiMo-V2-Flash December 16, 2025
- MiMo-V2-Pro March 18, 2026
- MiMo-V2.5 April 23, 2026
- MiMo-V2.5-Pro April 23, 2026
- MiMo-V2.6-Pro September 22, 2026
- MiMo-V2.6-Flash September 22, 2026
Compare MiMo-V2-Flash with another model →
See MiMo-V2-Flash on the interactive timeline →