← All comparisons
Grok 4.20 0309 (Reasoning) vs Kimi K2.5 (Non-reasoning)
xAI vs Kimi — side-by-side benchmark comparison
| Grok 4.20 0309 (Reasoning) | Kimi K2.5 (Non-reasoning) | |
|---|---|---|
| Intelligence Index | 48.5 | 37.3 |
| Coding Index | 42.2 | 25.8 |
| Math Index | — | — |
| Output speed (tok/s) | 217.8 | 33.5 |
| Blended price ($/1M) | $3.00 | $1.20 |
| Time to first token (s) | 13.18s | 1.23s |
| aime | — | — |
| aime 25 | — | — |
| artificial analysis coding index | 42.20 | 25.80 |
| artificial analysis intelligence index | 48.50 | 37.30 |
| artificial analysis math index | — | — |
| gpqa | 88.5% | 78.9% |
| hle | 30.0% | 12.3% |
| ifbench | 82.9% | 43.7% |
| lcr | 59.0% | 59.0% |
| livecodebench | — | — |
| math 500 | — | — |
| mmlu pro | — | — |
| scicode | 44.7% | 39.6% |
| tau2 | 96.5% | 81.3% |
| terminalbench hard | 40.9% | 18.9% |
Benchmark data from Artificial Analysis.