Magistral Small
| Released | 2025-06-10 |
|---|---|
| Price per 1M | $0.50 in · $1.50 out |
| Overall score | 39.7 (early, few results) |
| Context window | 131,072 tokens |
| Provider | Mistral |
| Weights | Open weights (Apache 2.0) |
Scores
| Dimension | Score | Rank | Benchmarks |
|---|---|---|---|
| Overall: Balanced | 39.7 | 5 | |
| Overall: Coding | 38.3 | 4 | |
| Overall: Reasoning | 39.8 | 5 | |
| Overall: Agentic work | 38.3 | 4 | |
| Reasoning | 38.3 | 4 | |
| Math | 41.8 | 1 | |
| Value vs price | −7.3 |
Evidence
Reasoning 38.3 ±5.9
| Benchmark | Result | Source | Date |
|---|---|---|---|
| GPQA Diamond | 56.1% | Epoch AI | 2026-08-06 |
| ARC-AGI-2 | 0.0% | Epoch AI | 2026-11-11 |
| ARC-AGI-1 (ARC Prize verified) | 5.0% | ARC Prize | 2026-11-11 |
| ARC-AGI-2 (ARC Prize verified) | 0.0% | ARC Prize | 2026-11-11 |
| ARC-AGI-1 | 5.0% | Epoch AI | 2026-11-11 |
| Kagi Reasoning Benchmark | 6.3% | Kagi | 2026-11-11 |
Math 41.8 ±3.6
| Benchmark | Result | Source | Date |
|---|---|---|---|
| OTIS Mock AIME 2024-2025 | 30.0% | Epoch AI | 2026-08-06 |
Names matched to this model
magistral-small, magistral-small-2506