Mistral Large 4 Preview
- Publisher
- Mistral AI
- Input
- Output
- Released
- 6 Oct 2026
Cite
Notes
Only stored in your browser.
Price & speed via Artificial Analysis · ranked across tracked models
| Reasoning effort | Index | Speed |
|---|---|---|
| Default | 25.3 | 52 t/s |
| high | 26.1 | 65 t/s |
| medium | 23.1 | 61 t/s |
Intelligence Index per reasoning-effort setting, via Artificial Analysis · pricing identical across tiers
Reported on 9 evals across 7 domains - best 80.4% on MedScribe · #53 of 92