← All comparisons
Claude 4.1 Opus (Non-reasoning) vs Kimi K2 0905
Anthropic vs Kimi — side-by-side benchmark comparison
| Claude 4.1 Opus (Non-reasoning) | Kimi K2 0905 | |
|---|---|---|
| Intelligence Index | 36.0 | 30.9 |
| Coding Index | — | 25.9 |
| Math Index | — | 57.3 |
| Output speed (tok/s) | 44.7 | 24.6 |
| Blended price ($/1M) | $32.81 | $1.07 |
| Time to first token (s) | 1.63s | 1.44s |
| aime | — | — |
| aime 25 | — | 57.3% |
| artificial analysis coding index | — | 25.90 |
| artificial analysis intelligence index | 36.00 | 30.90 |
| artificial analysis math index | — | 57.30 |
| gpqa | — | 76.7% |
| hle | — | 6.3% |
| ifbench | — | 41.7% |
| lcr | — | 52.3% |
| livecodebench | — | 61.0% |
| math 500 | — | — |
| mmlu pro | — | 81.9% |
| scicode | — | 30.7% |
| tau2 | — | 73.4% |
| terminalbench hard | — | 23.5% |
Benchmark data from Artificial Analysis.