← All comparisons
Ministral 3 3B vs DeepSeek R1 Distill Qwen 1.5B
Mistral vs DeepSeek — side-by-side benchmark comparison
| Ministral 3 3B | DeepSeek R1 Distill Qwen 1.5B | |
|---|---|---|
| Intelligence Index | 11.2 | 9.1 |
| Coding Index | 4.8 | — |
| Math Index | 22.0 | 22.0 |
| Output speed (tok/s) | 200.8 | 0.0 |
| Blended price ($/1M) | $0.10 | $0.00 |
| Time to first token (s) | 0.34s | 0.00s |
| aime | — | 17.7% |
| aime 25 | 22.0% | 22.0% |
| artificial analysis coding index | 4.80 | — |
| artificial analysis intelligence index | 11.20 | 9.10 |
| artificial analysis math index | 22.00 | 22.00 |
| gpqa | 35.8% | 9.8% |
| hle | 5.3% | 3.3% |
| ifbench | 26.8% | 13.2% |
| lcr | 11.7% | 0.3% |
| livecodebench | 24.7% | 7.0% |
| math 500 | — | 68.7% |
| mmlu pro | 52.4% | 26.9% |
| scicode | 14.4% | 6.6% |
| tau2 | 24.9% | — |
| terminalbench hard | 0.0% | — |
Benchmark data from Artificial Analysis.