VISION & MULTIMODAL · 紅葉
Mistral
7 models from Mistral, mistralai, living in Maple Highlands.
Visit the Mistral village in 3D
| Model | Context | Input | Output | Accepts |
|---|---|---|---|---|
| Mistral Large 3 2512mistralai/mistral-large-2512 | 262K | $0.500 | $1.50 | text, files, images |
| Mistral Medium 3.1mistralai/mistral-medium-3.1 | 131K | $0.400 | $2.00 | text, files, images |
| Mistral Medium 3mistralai/mistral-medium-3 | 131K | $0.400 | $2.00 | text, files, images |
| Sabamistralai/mistral-saba | 33K | $0.200 | $0.600 | text, files |
| Mistral Large 2407mistralai/mistral-large-2407 | 131K | $2.00 | $6.00 | text, files |
| Mixtral 8x22B Instructmistralai/mixtral-8x22b-instruct | 66K | $2.00 | $6.00 | text, files |
| Mistral Largemistralai/mistral-large | 128K | $2.00 | $6.00 | text, files |
Questions and answers
Which Mistral model has the largest context window?
Mistral Large 3 2512 (Mistral), with 262K tokens.
Which Mistral model is cheapest per input token?
Saba, at $0.200 per million input tokens ($0.600 per million output tokens).
Which Mistral models accept images?
Mistral Large 3 2512, Mistral Medium 3.1, Mistral Medium 3 (3 of 7).
Which Mistral models support reasoning?
None in the 2026-10-24 snapshot.
Answers are computed from the 2026-10-24 catalog snapshot.