← All comparisons
Claude 4.1 Opus (Non-reasoning) vs Claude Opus 4.5 (Reasoning)
Anthropic vs Anthropic — side-by-side benchmark comparison
| Claude 4.1 Opus (Non-reasoning) | Claude Opus 4.5 (Reasoning) | |
|---|---|---|
| Intelligence Index | 36.0 | 49.7 |
| Coding Index | — | 47.8 |
| Math Index | — | 91.3 |
| Output speed (tok/s) | 44.7 | 73.2 |
| Blended price ($/1M) | $32.81 | $10.94 |
| Time to first token (s) | 1.63s | 11.69s |
| aime | — | — |
| aime 25 | — | 91.3% |
| artificial analysis coding index | — | 47.80 |
| artificial analysis intelligence index | 36.00 | 49.70 |
| artificial analysis math index | — | 91.30 |
| gpqa | — | 86.6% |
| hle | — | 28.4% |
| ifbench | — | 58.0% |
| lcr | — | 74.0% |
| livecodebench | — | 87.1% |
| math 500 | — | — |
| mmlu pro | — | 89.5% |
| scicode | — | 49.5% |
| tau2 | — | 89.5% |
| terminalbench hard | — | 47.0% |
Benchmark data from Artificial Analysis.