← All comparisons
Devstral Small (Jul '25) vs Olmo 3 32B Think
Mistral vs Allen Institute for AI — side-by-side benchmark comparison
| Devstral Small (Jul '25) | Olmo 3 32B Think | |
|---|---|---|
| Intelligence Index | 15.2 | 12.1 |
| Coding Index | 12.1 | 10.5 |
| Math Index | 29.3 | 73.7 |
| Output speed (tok/s) | 183.4 | 0.0 |
| Blended price ($/1M) | $0.15 | $0.00 |
| Time to first token (s) | 0.40s | 0.00s |
| aime | 0.3% | — |
| aime 25 | 29.3% | 73.7% |
| artificial analysis coding index | 12.10 | 10.50 |
| artificial analysis intelligence index | 15.20 | 12.10 |
| artificial analysis math index | 29.30 | 73.70 |
| gpqa | 41.4% | 61.0% |
| hle | 3.7% | 5.9% |
| ifbench | 34.6% | 49.1% |
| lcr | 17.0% | 0.0% |
| livecodebench | 25.4% | 67.2% |
| math 500 | 63.5% | — |
| mmlu pro | 62.2% | 75.9% |
| scicode | 24.3% | 28.6% |
| tau2 | 28.4% | 0.0% |
| terminalbench hard | 6.1% | 1.5% |
Benchmark data from Artificial Analysis.