← All comparisons
Qwen3.5 122B A10B (Non-reasoning) vs Gemini 2.5 Flash (Reasoning)
Alibaba vs Google — side-by-side benchmark comparison
| Qwen3.5 122B A10B (Non-reasoning) | Gemini 2.5 Flash (Reasoning) | |
|---|---|---|
| Intelligence Index | 35.9 | 27.0 |
| Coding Index | 31.6 | 22.2 |
| Math Index | — | 73.3 |
| Output speed (tok/s) | 165.9 | 205.5 |
| Blended price ($/1M) | $1.10 | $0.85 |
| Time to first token (s) | 1.06s | 10.67s |
| aime | — | 82.3% |
| aime 25 | — | 73.3% |
| artificial analysis coding index | 31.60 | 22.20 |
| artificial analysis intelligence index | 35.90 | 27.00 |
| artificial analysis math index | — | 73.30 |
| gpqa | 82.7% | 79.0% |
| hle | 14.8% | 11.1% |
| ifbench | 50.8% | 50.3% |
| lcr | 56.0% | 61.7% |
| livecodebench | — | 69.5% |
| math 500 | — | 98.1% |
| mmlu pro | — | 83.2% |
| scicode | 35.6% | 39.4% |
| tau2 | 84.5% | 31.6% |
| terminalbench hard | 29.5% | 13.6% |
Benchmark data from Artificial Analysis.