← All comparisons
Qwen3.5 122B A10B (Non-reasoning) vs Claude Opus 4.5 (Reasoning)
Alibaba vs Anthropic — side-by-side benchmark comparison
| Qwen3.5 122B A10B (Non-reasoning) | Claude Opus 4.5 (Reasoning) | |
|---|---|---|
| Intelligence Index | 35.9 | 49.7 |
| Coding Index | 31.6 | 47.8 |
| Math Index | — | 91.3 |
| Output speed (tok/s) | 165.9 | 73.2 |
| Blended price ($/1M) | $1.10 | $10.94 |
| Time to first token (s) | 1.06s | 11.69s |
| aime | — | — |
| aime 25 | — | 91.3% |
| artificial analysis coding index | 31.60 | 47.80 |
| artificial analysis intelligence index | 35.90 | 49.70 |
| artificial analysis math index | — | 91.30 |
| gpqa | 82.7% | 86.6% |
| hle | 14.8% | 28.4% |
| ifbench | 50.8% | 58.0% |
| lcr | 56.0% | 74.0% |
| livecodebench | — | 87.1% |
| math 500 | — | — |
| mmlu pro | — | 89.5% |
| scicode | 35.6% | 49.5% |
| tau2 | 84.5% | 89.5% |
| terminalbench hard | 29.5% | 47.0% |
Benchmark data from Artificial Analysis.