← All comparisons
Magistral Medium 1.2 vs Claude 3.5 Sonnet (June '24)
Mistral vs Anthropic — side-by-side benchmark comparison
| Magistral Medium 1.2 | Claude 3.5 Sonnet (June '24) | |
|---|---|---|
| Intelligence Index | 27.1 | 14.2 |
| Coding Index | 21.7 | 26.0 |
| Math Index | 82.0 | — |
| Output speed (tok/s) | 39.4 | 0.0 |
| Blended price ($/1M) | $2.75 | $6.56 |
| Time to first token (s) | 0.54s | 0.00s |
| aime | — | 9.7% |
| aime 25 | 82.0% | — |
| artificial analysis coding index | 21.70 | 26.00 |
| artificial analysis intelligence index | 27.10 | 14.20 |
| artificial analysis math index | 82.00 | — |
| gpqa | 73.9% | 56.0% |
| hle | 9.6% | 3.7% |
| ifbench | 43.0% | — |
| lcr | 51.3% | — |
| livecodebench | 75.0% | — |
| math 500 | — | 69.5% |
| mmlu pro | 81.5% | 75.1% |
| scicode | 39.2% | 31.6% |
| tau2 | 52.0% | — |
| terminalbench hard | 12.9% | — |
Benchmark data from Artificial Analysis.