← All comparisons
Olmo 3 32B Think vs Qwen3 235B A22B (Non-reasoning)
Allen Institute for AI vs Alibaba — side-by-side benchmark comparison
| Olmo 3 32B Think | Qwen3 235B A22B (Non-reasoning) | |
|---|---|---|
| Intelligence Index | 12.1 | 17.0 |
| Coding Index | 10.5 | 14.0 |
| Math Index | 73.7 | 23.7 |
| Output speed (tok/s) | 0.0 | 60.8 |
| Blended price ($/1M) | $0.00 | $0.79 |
| Time to first token (s) | 0.00s | 1.30s |
| aime | — | 32.7% |
| aime 25 | 73.7% | 23.7% |
| artificial analysis coding index | 10.50 | 14.00 |
| artificial analysis intelligence index | 12.10 | 17.00 |
| artificial analysis math index | 73.70 | 23.70 |
| gpqa | 61.0% | 61.3% |
| hle | 5.9% | 4.7% |
| ifbench | 49.1% | 36.6% |
| lcr | 0.0% | 0.0% |
| livecodebench | 67.2% | 34.3% |
| math 500 | — | 90.2% |
| mmlu pro | 75.9% | 76.2% |
| scicode | 28.6% | 29.9% |
| tau2 | 0.0% | 27.2% |
| terminalbench hard | 1.5% | 6.1% |
Benchmark data from Artificial Analysis.