Head to head
Xiaomi MiMo V2.6 Pro vs DeepSeek V4.1 Flash
Which is better, Xiaomi MiMo V2.6 Pro or DeepSeek V4.1 Flash?
Xiaomi MiMo V2.6 Pro finishes ahead of DeepSeek V4.1 Flash on the index, 89.4 to 88.3. It is the stronger of the two on Craft, while DeepSeek V4.1 Flash still leads on Speed. Both are scored on the same rubric, in the same week, and neither placing is sponsored.
Updated
Xiaomi MiMo V2.6 Pro
06 of 73
89.4
The strongest large open-weights model on the independent index, under MIT, for under a dollar out.
DeepSeek V4.1 Flash
07 of 73
88.3
Frontier-adjacent answers for thirty cents per million input tokens, with the weights under MIT.
The four scores, side by side
| Score | Xiaomi MiMo V2.6 Pro | DeepSeek V4.1 Flash | Difference |
|---|---|---|---|
| Craft | 88 | 84 | +4 |
| Speed | 80 | 88 | +8 |
| Control | 93 | 94 | +1 |
| Value | 95 | 96 | +1 |
| Index score | 89.4 | 88.3 | +1.1 |
Which to pick
Pick Xiaomi MiMo V2.6 Pro unless Speed is what decides it for you, which is exactly where DeepSeek V4.1 Flash is the better answer.
Xiaomi MiMo V2.6 Pro
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
DeepSeek V4.1 Flash
V4.1 Flash is the first model in DeepSeek's new architecture family, released on 10 September, and it replaced V4 Flash outright on the API. It scores above V4 Pro on both independent leaderboards while costing about a quarter as much, takes a million tokens of context and reads images. The weights are published under MIT. For high-volume work where the bill is the constraint, it is one of the best bargains in the index.
Where it shines
- Beats DeepSeek's own V4 Pro at about a quarter of the price
- MIT licence on the published weights
- Million-token context and 384,000 tokens of output