← All comparisons
Llama 4 Scout vs Mistral Small 3.1
Meta vs Mistral — side-by-side benchmark comparison
| Llama 4 Scout | Mistral Small 3.1 | |
|---|---|---|
| Intelligence Index | 13.5 | 14.5 |
| Coding Index | 6.7 | 13.9 |
| Math Index | 14.0 | 3.7 |
| Output speed (tok/s) | 113.8 | 163.2 |
| Blended price ($/1M) | $0.29 | $0.14 |
| Time to first token (s) | 0.56s | 0.49s |
| aime | 28.3% | 9.3% |
| aime 25 | 14.0% | 3.7% |
| artificial analysis coding index | 6.70 | 13.90 |
| artificial analysis intelligence index | 13.50 | 14.50 |
| artificial analysis math index | 14.00 | 3.70 |
| gpqa | 58.7% | 45.4% |
| hle | 4.3% | 4.8% |
| ifbench | 39.5% | 29.9% |
| lcr | 25.8% | 19.7% |
| livecodebench | 29.9% | 21.2% |
| math 500 | 84.4% | 70.7% |
| mmlu pro | 75.2% | 65.9% |
| scicode | 17.0% | 26.5% |
| tau2 | 15.5% | 25.1% |
| terminalbench hard | 1.5% | 7.6% |
Benchmark data from Artificial Analysis.