← All comparisons
Gemini 3.1 Pro Preview vs Qwen3.5 122B A10B (Non-reasoning)
Google vs Alibaba — side-by-side benchmark comparison
| Gemini 3.1 Pro Preview | Qwen3.5 122B A10B (Non-reasoning) | |
|---|---|---|
| Intelligence Index | 57.2 | 35.9 |
| Coding Index | 55.5 | 31.6 |
| Math Index | — | — |
| Output speed (tok/s) | 136.8 | 165.9 |
| Blended price ($/1M) | $4.50 | $1.10 |
| Time to first token (s) | 25.68s | 1.06s |
| aime | — | — |
| aime 25 | — | — |
| artificial analysis coding index | 55.50 | 31.60 |
| artificial analysis intelligence index | 57.20 | 35.90 |
| artificial analysis math index | — | — |
| gpqa | 94.1% | 82.7% |
| hle | 44.7% | 14.8% |
| ifbench | 77.1% | 50.8% |
| lcr | 72.7% | 56.0% |
| livecodebench | — | — |
| math 500 | — | — |
| mmlu pro | — | — |
| scicode | 58.9% | 35.6% |
| tau2 | 95.6% | 84.5% |
| terminalbench hard | 53.8% | 29.5% |
Benchmark data from Artificial Analysis.