← All comparisons
Devstral Small 2 vs Claude 4.1 Opus (Reasoning)
Mistral vs Anthropic — side-by-side benchmark comparison
| Devstral Small 2 | Claude 4.1 Opus (Reasoning) | |
|---|---|---|
| Intelligence Index | 19.5 | 42.0 |
| Coding Index | 20.7 | 36.5 |
| Math Index | 34.3 | 80.3 |
| Output speed (tok/s) | 75.2 | 44.5 |
| Blended price ($/1M) | $0.00 | $32.81 |
| Time to first token (s) | 0.73s | 8.55s |
| aime | — | — |
| aime 25 | 34.3% | 80.3% |
| artificial analysis coding index | 20.70 | 36.50 |
| artificial analysis intelligence index | 19.50 | 42.00 |
| artificial analysis math index | 34.30 | 80.30 |
| gpqa | 53.2% | 80.9% |
| hle | 3.4% | 11.9% |
| ifbench | 31.2% | 55.4% |
| lcr | 24.0% | 66.3% |
| livecodebench | 34.8% | 65.4% |
| math 500 | — | — |
| mmlu pro | 67.8% | 88.0% |
| scicode | 28.8% | 40.9% |
| tau2 | 23.4% | 71.4% |
| terminalbench hard | 16.7% | 34.3% |
Benchmark data from Artificial Analysis.