← All comparisons
Magistral Medium 1.2 vs GPT-5 mini (high)
Mistral vs OpenAI — side-by-side benchmark comparison
| Magistral Medium 1.2 | GPT-5 mini (high) | |
|---|---|---|
| Intelligence Index | 27.1 | 41.2 |
| Coding Index | 21.7 | 35.3 |
| Math Index | 82.0 | 90.7 |
| Output speed (tok/s) | 39.4 | 96.1 |
| Blended price ($/1M) | $2.75 | $0.69 |
| Time to first token (s) | 0.54s | 77.65s |
| aime | — | — |
| aime 25 | 82.0% | 90.7% |
| artificial analysis coding index | 21.70 | 35.30 |
| artificial analysis intelligence index | 27.10 | 41.20 |
| artificial analysis math index | 82.00 | 90.70 |
| gpqa | 73.9% | 82.8% |
| hle | 9.6% | 19.7% |
| ifbench | 43.0% | 75.4% |
| lcr | 51.3% | 68.0% |
| livecodebench | 75.0% | 83.8% |
| math 500 | — | — |
| mmlu pro | 81.5% | 83.7% |
| scicode | 39.2% | 39.2% |
| tau2 | 52.0% | 68.4% |
| terminalbench hard | 12.9% | 33.3% |
Benchmark data from Artificial Analysis.