← All comparisons
Claude 3 Haiku vs Qwen3 Max Thinking (Preview)
Anthropic vs Alibaba — side-by-side benchmark comparison
| Claude 3 Haiku | Qwen3 Max Thinking (Preview) | |
|---|---|---|
| Intelligence Index | 12.3 | 32.5 |
| Coding Index | 6.7 | 24.5 |
| Math Index | — | 82.3 |
| Output speed (tok/s) | 0.0 | 49.3 |
| Blended price ($/1M) | $0.50 | $2.40 |
| Time to first token (s) | 0.00s | 1.92s |
| aime | 1.0% | — |
| aime 25 | — | 82.3% |
| artificial analysis coding index | 6.70 | 24.50 |
| artificial analysis intelligence index | 12.30 | 32.50 |
| artificial analysis math index | — | 82.30 |
| gpqa | 37.4% | 77.6% |
| hle | 3.9% | 12.0% |
| ifbench | 36.1% | 53.8% |
| lcr | 21.0% | 57.7% |
| livecodebench | 15.4% | 53.5% |
| math 500 | 39.4% | — |
| mmlu pro | — | 82.4% |
| scicode | 18.6% | 38.7% |
| tau2 | 21.1% | 83.6% |
| terminalbench hard | 0.8% | 17.4% |
Benchmark data from Artificial Analysis.