← All comparisons
Claude 3.7 Sonnet (Non-reasoning) vs Qwen3 4B 2507 Instruct
Anthropic vs Alibaba — side-by-side benchmark comparison
| Claude 3.7 Sonnet (Non-reasoning) | Qwen3 4B 2507 Instruct | |
|---|---|---|
| Intelligence Index | 30.8 | 12.9 |
| Coding Index | 26.7 | 9.0 |
| Math Index | 21.0 | 52.3 |
| Output speed (tok/s) | 0.0 | 0.0 |
| Blended price ($/1M) | $6.56 | $0.00 |
| Time to first token (s) | 0.00s | 0.00s |
| aime | 22.3% | — |
| aime 25 | 21.0% | 52.3% |
| artificial analysis coding index | 26.70 | 9.00 |
| artificial analysis intelligence index | 30.80 | 12.90 |
| artificial analysis math index | 21.00 | 52.30 |
| gpqa | 65.6% | 51.7% |
| hle | 4.8% | 4.7% |
| ifbench | 44.0% | 33.5% |
| lcr | 48.3% | 7.3% |
| livecodebench | 39.4% | 37.7% |
| math 500 | 85.0% | — |
| mmlu pro | 80.3% | 67.2% |
| scicode | 37.6% | 18.1% |
| tau2 | 50.0% | 26.6% |
| terminalbench hard | 21.2% | 4.5% |
Benchmark data from Artificial Analysis.