← All comparisons
Olmo 3.1 32B Think vs KAT-Coder-Pro V1
Allen Institute for AI vs KwaiKAT — side-by-side benchmark comparison
| Olmo 3.1 32B Think | KAT-Coder-Pro V1 | |
|---|---|---|
| Intelligence Index | 13.9 | 36.0 |
| Coding Index | 9.8 | 18.3 |
| Math Index | 77.3 | 94.7 |
| Output speed (tok/s) | 0.0 | 108.8 |
| Blended price ($/1M) | $0.00 | $0.53 |
| Time to first token (s) | 0.00s | 1.40s |
| aime | — | — |
| aime 25 | 77.3% | 94.7% |
| artificial analysis coding index | 9.80 | 18.30 |
| artificial analysis intelligence index | 13.90 | 36.00 |
| artificial analysis math index | 77.30 | 94.70 |
| gpqa | 59.1% | 76.4% |
| hle | 6.0% | 33.4% |
| ifbench | 66.0% | 68.4% |
| lcr | 0.0% | 74.0% |
| livecodebench | 69.5% | 74.7% |
| math 500 | — | — |
| mmlu pro | 76.3% | 81.3% |
| scicode | 29.3% | 36.6% |
| tau2 | 0.0% | 88.6% |
| terminalbench hard | 0.0% | 9.1% |
Benchmark data from Artificial Analysis.