← All comparisons
Qwen3.5 2B (Non-reasoning) vs o4-mini (high)
Alibaba vs OpenAI — side-by-side benchmark comparison
| Qwen3.5 2B (Non-reasoning) | o4-mini (high) | |
|---|---|---|
| Intelligence Index | 14.7 | 33.1 |
| Coding Index | 4.9 | 25.6 |
| Math Index | — | 90.7 |
| Output speed (tok/s) | 272.0 | 160.5 |
| Blended price ($/1M) | $0.04 | $1.93 |
| Time to first token (s) | 0.27s | 23.07s |
| aime | — | 94.0% |
| aime 25 | — | 90.7% |
| artificial analysis coding index | 4.90 | 25.60 |
| artificial analysis intelligence index | 14.70 | 33.10 |
| artificial analysis math index | — | 90.70 |
| gpqa | 43.8% | 78.4% |
| hle | 4.9% | 17.5% |
| ifbench | 29.1% | 68.7% |
| lcr | 13.7% | 55.0% |
| livecodebench | — | 85.9% |
| math 500 | — | 98.9% |
| mmlu pro | — | 83.2% |
| scicode | 7.2% | 46.5% |
| tau2 | 81.6% | 55.6% |
| terminalbench hard | 3.8% | 15.2% |
Benchmark data from Artificial Analysis.