← All comparisons
Claude 4 Sonnet (Non-reasoning) vs Claude 4.1 Opus (Reasoning)
Anthropic vs Anthropic — side-by-side benchmark comparison
| Claude 4 Sonnet (Non-reasoning) | Claude 4.1 Opus (Reasoning) | |
|---|---|---|
| Intelligence Index | 33.0 | 42.0 |
| Coding Index | 30.6 | 36.5 |
| Math Index | 38.0 | 80.3 |
| Output speed (tok/s) | 53.9 | 44.5 |
| Blended price ($/1M) | $6.56 | $32.81 |
| Time to first token (s) | 0.82s | 8.55s |
| aime | 40.7% | — |
| aime 25 | 38.0% | 80.3% |
| artificial analysis coding index | 30.60 | 36.50 |
| artificial analysis intelligence index | 33.00 | 42.00 |
| artificial analysis math index | 38.00 | 80.30 |
| gpqa | 68.3% | 80.9% |
| hle | 4.0% | 11.9% |
| ifbench | 45.4% | 55.4% |
| lcr | 44.3% | 66.3% |
| livecodebench | 44.9% | 65.4% |
| math 500 | 93.4% | — |
| mmlu pro | 83.7% | 88.0% |
| scicode | 37.3% | 40.9% |
| tau2 | 52.3% | 71.4% |
| terminalbench hard | 27.3% | 34.3% |
Benchmark data from Artificial Analysis.