← All comparisons
Claude 4.5 Haiku (Reasoning) vs GPT-5.1 Codex mini (high)
Anthropic vs OpenAI — side-by-side benchmark comparison
| Claude 4.5 Haiku (Reasoning) | GPT-5.1 Codex mini (high) | |
|---|---|---|
| Intelligence Index | 37.1 | 38.6 |
| Coding Index | 32.6 | 36.4 |
| Math Index | 83.7 | 91.7 |
| Output speed (tok/s) | 142.2 | 218.5 |
| Blended price ($/1M) | $2.19 | $0.69 |
| Time to first token (s) | 10.48s | 3.35s |
| aime | — | — |
| aime 25 | 83.7% | 91.7% |
| artificial analysis coding index | 32.60 | 36.40 |
| artificial analysis intelligence index | 37.10 | 38.60 |
| artificial analysis math index | 83.70 | 91.70 |
| gpqa | 67.2% | 81.3% |
| hle | 9.7% | 16.9% |
| ifbench | 54.3% | 67.9% |
| lcr | 70.3% | 62.7% |
| livecodebench | 61.5% | 83.6% |
| math 500 | — | — |
| mmlu pro | 76.0% | 82.0% |
| scicode | 43.3% | 42.6% |
| tau2 | 54.7% | 62.9% |
| terminalbench hard | 27.3% | 33.3% |
Benchmark data from Artificial Analysis.