← All comparisons
Mistral Small 3.2 vs Qwen3 235B A22B 2507 (Reasoning)
Mistral vs Alibaba — side-by-side benchmark comparison
| Mistral Small 3.2 | Qwen3 235B A22B 2507 (Reasoning) | |
|---|---|---|
| Intelligence Index | 15.1 | 29.5 |
| Coding Index | 13.3 | 23.2 |
| Math Index | 27.0 | 91.0 |
| Output speed (tok/s) | 133.0 | 62.5 |
| Blended price ($/1M) | $0.13 | $0.84 |
| Time to first token (s) | 0.36s | 1.21s |
| aime | 32.3% | 94.0% |
| aime 25 | 27.0% | 91.0% |
| artificial analysis coding index | 13.30 | 23.20 |
| artificial analysis intelligence index | 15.10 | 29.50 |
| artificial analysis math index | 27.00 | 91.00 |
| gpqa | 50.5% | 79.0% |
| hle | 4.3% | 15.0% |
| ifbench | 33.5% | 51.2% |
| lcr | 17.3% | 67.0% |
| livecodebench | 27.5% | 78.8% |
| math 500 | 88.3% | 98.4% |
| mmlu pro | 68.1% | 84.3% |
| scicode | 26.4% | 42.4% |
| tau2 | 29.5% | 53.2% |
| terminalbench hard | 6.8% | 13.6% |
Benchmark data from Artificial Analysis.