← All comparisons
Olmo 3.1 32B Instruct vs Kimi K2 Thinking
Allen Institute for AI vs Kimi — side-by-side benchmark comparison
| Olmo 3.1 32B Instruct | Kimi K2 Thinking | |
|---|---|---|
| Intelligence Index | 12.2 | 40.9 |
| Coding Index | 5.6 | 34.8 |
| Math Index | — | 94.7 |
| Output speed (tok/s) | 0.0 | 130.3 |
| Blended price ($/1M) | $0.00 | $1.07 |
| Time to first token (s) | 0.00s | 0.84s |
| aime | — | — |
| aime 25 | — | 94.7% |
| artificial analysis coding index | 5.60 | 34.80 |
| artificial analysis intelligence index | 12.20 | 40.90 |
| artificial analysis math index | — | 94.70 |
| gpqa | 53.9% | 83.8% |
| hle | 4.9% | 22.3% |
| ifbench | 39.2% | 68.1% |
| lcr | 0.0% | 66.3% |
| livecodebench | — | 85.3% |
| math 500 | — | — |
| mmlu pro | — | 84.8% |
| scicode | 16.7% | 42.4% |
| tau2 | 21.3% | 93.0% |
| terminalbench hard | 0.0% | 31.1% |
Benchmark data from Artificial Analysis.