MiMo-V2.5
Xiaomi · Undisclosed (MoE)
The breakout of 2026
1M context, omni-modal understanding and the #1 call volume on OpenRouter in July 2026. Improving fast on Chinese and video, but public eval samples remain thin — evidence grade B.
Last updated Oct 18, 2026Editorial scoring
Bottom line
Eight-dimension breakdown
editorial mapping · 0–100- MiMo-V2.590.1
Weights: perception 20 / documents 18 / charts 14 / video 14 / agents 14 / generation 6 / Chinese 6 / ecosystem 8. Definitions and sources on the methodology page.
Public benchmarks
as reported, not converted| Benchmark / metric | Value | Note |
|---|---|---|
| OpenRouter 调用量 | #1 | July 2026 |
| 上下文 | 1M | Official spec |
| 调用量增长 | +616% | May–Jul 2026 |
Public scores come from different versions and prompts and are not interchangeable with our composite; they are one evidence input.
Scenario mapping
five stress tests, one scale| Scenario | Score | Observation |
|---|---|---|
| Chinese receipt OCR | 88 · Fine on standard receipts | |
| Chart data extraction | 88 · Stable on simple charts | |
| 30-min video QA | 93 · Best value for long video | |
| Mobile GUI control | 89 · Well tuned to own devices | |
| Chinese cultural context | 89 · Strong everyday scenes; average academic context |
Strengths and limitations
Strengths
- Half the flagship price with standout video value.
- Rich device/edge data; steadier on real-world photos.
- 1M context is friendly to long video and many-image inputs.
Limitations
- Sparse public evaluations; evidence grade capped at B.
- Enterprise SLAs and on-prem options are still maturing.
- Complex document layout trails document specialists.
Price and real cost
List price
¥6/M tokens
≈ ¥6 / 1M output tokens. For 1,000 images at 1,024px, ~1,100 image tokens each and 300 output tokens, the monthly cost is about ¥3.6 (assuming input at 25% of output price).
Use your volumeAbout MiMo-V2.5
Is MiMo-V2.5 open source?
Licence: Closed API + partial open weights. No public weights; access is via vendor API or an on-prem licence.
How much does MiMo-V2.5 cost?
≈ ¥6 / 1M output tokens. It ranks #7 with a composite of 90.1. Real bills depend on resolution, tiling, caching and volume discounts.
What is MiMo-V2.5 best for?
Edge and in-car multimodal interaction; Short-video understanding and tagging; Cost-sensitive batch image tasks。
Why does MiMo-V2.5 score 90.1?
The composite weights eight dimensions: perception 20%, documents 18%, charts 14%, video 14%, agents 14%, generation 6%, Chinese 6%, ecosystem 8%. Its strongest dimension is Video & long-horizon reasoning; its weakest is Generation & editing.
Personal ranking · Not an official score
Every ranking, score and comparison on this page is an independent personal ranking, not an official score, and represents no vendor, institution or regulator. Scores are a subjective mapping of public evidence under our own weights — change the weights and the order changes. Disagree, or spotted a factual error? Contact the site owner: jizhi2024@foxmail.com — we check each report and log corrections in the changelog.