← All comparisons
Pixtral Large vs Qwen3 VL 235B A22B (Reasoning)
Mistral vs Alibaba — side-by-side benchmark comparison
| Pixtral Large | Qwen3 VL 235B A22B (Reasoning) | |
|---|---|---|
| Intelligence Index | 14.0 | 27.6 |
| Coding Index | — | 20.9 |
| Math Index | 2.3 | 88.3 |
| Output speed (tok/s) | 54.9 | 35.6 |
| Blended price ($/1M) | $3.00 | $2.17 |
| Time to first token (s) | 0.50s | 5.14s |
| aime | 7.0% | — |
| aime 25 | 2.3% | 88.3% |
| artificial analysis coding index | — | 20.90 |
| artificial analysis intelligence index | 14.00 | 27.60 |
| artificial analysis math index | 2.30 | 88.30 |
| gpqa | 50.5% | 77.2% |
| hle | 3.6% | 10.1% |
| ifbench | 34.5% | 56.5% |
| lcr | 10.3% | 58.7% |
| livecodebench | 26.1% | 64.6% |
| math 500 | 71.4% | — |
| mmlu pro | 70.1% | 83.6% |
| scicode | 29.2% | 39.9% |
| tau2 | 36.5% | 54.1% |
| terminalbench hard | — | 11.4% |
Benchmark data from Artificial Analysis.