← All comparisons
GPT-5.1 (high) vs Claude 4.1 Opus (Non-reasoning)
OpenAI vs Anthropic — side-by-side benchmark comparison
| GPT-5.1 (high) | Claude 4.1 Opus (Non-reasoning) | |
|---|---|---|
| Intelligence Index | 47.7 | 36.0 |
| Coding Index | 44.7 | — |
| Math Index | 94.0 | — |
| Output speed (tok/s) | 129.6 | 44.7 |
| Blended price ($/1M) | $3.44 | $32.81 |
| Time to first token (s) | 19.36s | 1.63s |
| aime | — | — |
| aime 25 | 94.0% | — |
| artificial analysis coding index | 44.70 | — |
| artificial analysis intelligence index | 47.70 | 36.00 |
| artificial analysis math index | 94.00 | — |
| gpqa | 87.3% | — |
| hle | 26.5% | — |
| ifbench | 72.9% | — |
| lcr | 75.0% | — |
| livecodebench | 86.8% | — |
| math 500 | — | — |
| mmlu pro | 87.0% | — |
| scicode | 43.3% | — |
| tau2 | 81.9% | — |
| terminalbench hard | 45.5% | — |
Benchmark data from Artificial Analysis.