← All comparisons
Kimi K2 Thinking vs Llama 3.1 Tulu3 405B
Kimi vs Allen Institute for AI — side-by-side benchmark comparison
| Kimi K2 Thinking | Llama 3.1 Tulu3 405B | |
|---|---|---|
| Intelligence Index | 40.9 | 14.1 |
| Coding Index | 34.8 | — |
| Math Index | 94.7 | — |
| Output speed (tok/s) | 130.3 | 0.0 |
| Blended price ($/1M) | $1.07 | $0.00 |
| Time to first token (s) | 0.84s | 0.00s |
| aime | — | 13.3% |
| aime 25 | 94.7% | — |
| artificial analysis coding index | 34.80 | — |
| artificial analysis intelligence index | 40.90 | 14.10 |
| artificial analysis math index | 94.70 | — |
| gpqa | 83.8% | 51.6% |
| hle | 22.3% | 3.5% |
| ifbench | 68.1% | — |
| lcr | 66.3% | — |
| livecodebench | 85.3% | 29.1% |
| math 500 | — | 77.8% |
| mmlu pro | 84.8% | 71.6% |
| scicode | 42.4% | 30.2% |
| tau2 | 93.0% | — |
| terminalbench hard | 31.1% | — |
Benchmark data from Artificial Analysis.