← All comparisons
Kimi K2.6 (Non-reasoning) vs o3-mini (high)
Kimi vs OpenAI — side-by-side benchmark comparison
| Kimi K2.6 (Non-reasoning) | o3-mini (high) | |
|---|---|---|
| Intelligence Index | 42.9 | 25.2 |
| Coding Index | 38.4 | 17.3 |
| Math Index | — | — |
| Output speed (tok/s) | 39.0 | 187.0 |
| Blended price ($/1M) | $1.71 | $1.93 |
| Time to first token (s) | 1.28s | 15.73s |
| aime | — | 86.0% |
| aime 25 | — | — |
| artificial analysis coding index | 38.40 | 17.30 |
| artificial analysis intelligence index | 42.90 | 25.20 |
| artificial analysis math index | — | — |
| gpqa | 78.8% | 77.3% |
| hle | 18.2% | 12.3% |
| ifbench | 44.3% | 67.1% |
| lcr | 57.7% | 39.3% |
| livecodebench | — | 73.4% |
| math 500 | — | 98.5% |
| mmlu pro | — | 80.2% |
| scicode | 39.5% | 39.8% |
| tau2 | 93.9% | 31.3% |
| terminalbench hard | 37.9% | 6.1% |
Benchmark data from Artificial Analysis.