Mistral AI·released 2025-06-10
Scored on 4 benchmarks. 50% of its scores are independently reproduced — the rest are self-reported or unverified.
Every score, ranked, colored by evidence status.
| Benchmark | Score | Evidence status | Measured |
|---|---|---|---|
| GPQA diamond | 31.23accuracy (%) | reproduced· optimized | 2025-06-10 |
| OTIS Mock AIME 2024-2025 | 29.93accuracy (%) | reproduced· optimized | 2025-06-10 |
| ARC-AGI | 5accuracy (%) | unverified· optimized | 2025-06-10 |
| ARC-AGI-2 | 0accuracy (%) | unverified | 2025-06-10 |