← All comparisons
Qwen3.5 397B A17B (Non-reasoning) vs MiniMax-M2.1
Alibaba vs MiniMax — side-by-side benchmark comparison
| Qwen3.5 397B A17B (Non-reasoning) | MiniMax-M2.1 | |
|---|---|---|
| Intelligence Index | 40.1 | 39.4 |
| Coding Index | 37.4 | 32.8 |
| Math Index | — | 82.7 |
| Output speed (tok/s) | 53.5 | 188.8 |
| Blended price ($/1M) | $1.35 | $0.53 |
| Time to first token (s) | 1.85s | 2.43s |
| aime | — | — |
| aime 25 | — | 82.7% |
| artificial analysis coding index | 37.40 | 32.80 |
| artificial analysis intelligence index | 40.10 | 39.40 |
| artificial analysis math index | — | 82.70 |
| gpqa | 86.1% | 83.0% |
| hle | 18.8% | 22.2% |
| ifbench | 51.6% | 69.9% |
| lcr | 58.0% | 59.0% |
| livecodebench | — | 81.0% |
| math 500 | — | — |
| mmlu pro | — | 87.5% |
| scicode | 41.1% | 40.7% |
| tau2 | 83.9% | 85.4% |
| terminalbench hard | 35.6% | 28.8% |
Benchmark data from Artificial Analysis.