← All comparisons
GPT-5 Codex (high) vs Olmo 3 32B Think
OpenAI vs Allen Institute for AI — side-by-side benchmark comparison
| GPT-5 Codex (high) | Olmo 3 32B Think | |
|---|---|---|
| Intelligence Index | 44.6 | 12.1 |
| Coding Index | 38.9 | 10.5 |
| Math Index | 98.7 | 73.7 |
| Output speed (tok/s) | 173.2 | 0.0 |
| Blended price ($/1M) | $3.44 | $0.00 |
| Time to first token (s) | 7.64s | 0.00s |
| aime | — | — |
| aime 25 | 98.7% | 73.7% |
| artificial analysis coding index | 38.90 | 10.50 |
| artificial analysis intelligence index | 44.60 | 12.10 |
| artificial analysis math index | 98.70 | 73.70 |
| gpqa | 83.7% | 61.0% |
| hle | 25.6% | 5.9% |
| ifbench | 74.1% | 49.1% |
| lcr | 69.0% | 0.0% |
| livecodebench | 84.0% | 67.2% |
| math 500 | — | — |
| mmlu pro | 86.5% | 75.9% |
| scicode | 40.9% | 28.6% |
| tau2 | 86.8% | 0.0% |
| terminalbench hard | 37.9% | 1.5% |
Benchmark data from Artificial Analysis.