← All comparisons
gpt-oss-120b (high) vs Qwen3.5 397B A17B (Non-reasoning)
OpenAI vs Alibaba — side-by-side benchmark comparison
| gpt-oss-120b (high) | Qwen3.5 397B A17B (Non-reasoning) | |
|---|---|---|
| Intelligence Index | 33.3 | 40.1 |
| Coding Index | 28.6 | 37.4 |
| Math Index | 93.4 | — |
| Output speed (tok/s) | 356.8 | 53.5 |
| Blended price ($/1M) | $0.26 | $1.35 |
| Time to first token (s) | 0.51s | 1.85s |
| aime | — | — |
| aime 25 | 93.4% | — |
| artificial analysis coding index | 28.60 | 37.40 |
| artificial analysis intelligence index | 33.30 | 40.10 |
| artificial analysis math index | 93.40 | — |
| gpqa | 78.2% | 86.1% |
| hle | 18.5% | 18.8% |
| ifbench | 69.0% | 51.6% |
| lcr | 50.7% | 58.0% |
| livecodebench | 87.8% | — |
| math 500 | — | — |
| mmlu pro | 80.8% | — |
| scicode | 38.9% | 41.1% |
| tau2 | 65.8% | 83.9% |
| terminalbench hard | 23.5% | 35.6% |
Benchmark data from Artificial Analysis.