← All comparisons
GPT-4.1 vs Qwen3.5 27B (Non-reasoning)
OpenAI vs Alibaba — side-by-side benchmark comparison
| GPT-4.1 | Qwen3.5 27B (Non-reasoning) | |
|---|---|---|
| Intelligence Index | 26.3 | 37.2 |
| Coding Index | 21.8 | 33.4 |
| Math Index | 34.7 | — |
| Output speed (tok/s) | 137.8 | 95.3 |
| Blended price ($/1M) | $3.50 | $0.88 |
| Time to first token (s) | 0.58s | 1.40s |
| aime | 43.7% | — |
| aime 25 | 34.7% | — |
| artificial analysis coding index | 21.80 | 33.40 |
| artificial analysis intelligence index | 26.30 | 37.20 |
| artificial analysis math index | 34.70 | — |
| gpqa | 66.6% | 84.2% |
| hle | 4.6% | 13.2% |
| ifbench | 43.0% | 46.9% |
| lcr | 61.0% | 55.7% |
| livecodebench | 45.7% | — |
| math 500 | 91.3% | — |
| mmlu pro | 80.6% | — |
| scicode | 38.1% | 36.7% |
| tau2 | 47.1% | 87.1% |
| terminalbench hard | 13.6% | 31.8% |
Benchmark data from Artificial Analysis.