← All comparisons
Ministral 3 14B vs Qwen3 Omni 30B A3B (Reasoning)
Mistral vs Alibaba — side-by-side benchmark comparison
| Ministral 3 14B | Qwen3 Omni 30B A3B (Reasoning) | |
|---|---|---|
| Intelligence Index | 16.0 | 15.6 |
| Coding Index | 10.9 | 12.7 |
| Math Index | 30.0 | 74.0 |
| Output speed (tok/s) | 90.6 | 100.1 |
| Blended price ($/1M) | $0.20 | $0.43 |
| Time to first token (s) | 0.37s | 0.97s |
| aime | — | — |
| aime 25 | 30.0% | 74.0% |
| artificial analysis coding index | 10.90 | 12.70 |
| artificial analysis intelligence index | 16.00 | 15.60 |
| artificial analysis math index | 30.00 | 74.00 |
| gpqa | 57.2% | 72.6% |
| hle | 4.6% | 7.3% |
| ifbench | 32.0% | 43.4% |
| lcr | 22.0% | 0.0% |
| livecodebench | 35.1% | 67.9% |
| math 500 | — | — |
| mmlu pro | 69.3% | 79.2% |
| scicode | 23.6% | 30.6% |
| tau2 | 27.2% | 21.3% |
| terminalbench hard | 4.5% | 3.8% |
Benchmark data from Artificial Analysis.