Pyyan / Compare / Sarvam vs MiMo-V2.6-Flash
Open-Weight Models · verified 29 Sept 2026
| Specification | Sarvam | MiMo-V2.6-Flash |
|---|---|---|
| Summary | India-first model for Indic languages. | The small one, and still multimodal. |
| Context | — | 1M |
| Input ($/Mtok) | — | Weights only |
| Output ($/Mtok) | — | Weights only |
| Licence | Open weights | MIT |
| Max output | — | Not published |
| Parameters | — | 309B total, 15B active |
| Released | — | 21 Sep 2026 |
| Model ID | — | XiaomiMiMo/MiMo-V2.6-Flash |
| Category | Open-Weight Models | Open-Weight Models |
| Official | Sarvam AI ↗ | Xiaomi ↗ |
Highlighted rows are where these differ.
Best for multimodal work on your own cluster.