← All comparisons
Llama 3.1 Nemotron Instruct 70B vs Kimi K2 0905
NVIDIA vs Kimi — side-by-side benchmark comparison
| Llama 3.1 Nemotron Instruct 70B | Kimi K2 0905 | |
|---|---|---|
| Intelligence Index | 13.4 | 30.9 |
| Coding Index | 10.8 | 25.9 |
| Math Index | 11.0 | 57.3 |
| Output speed (tok/s) | 292.9 | 24.6 |
| Blended price ($/1M) | $1.20 | $1.07 |
| Time to first token (s) | 0.24s | 1.44s |
| aime | 24.7% | — |
| aime 25 | 11.0% | 57.3% |
| artificial analysis coding index | 10.80 | 25.90 |
| artificial analysis intelligence index | 13.40 | 30.90 |
| artificial analysis math index | 11.00 | 57.30 |
| gpqa | 46.5% | 76.7% |
| hle | 4.6% | 6.3% |
| ifbench | 30.7% | 41.7% |
| lcr | 7.0% | 52.3% |
| livecodebench | 16.9% | 61.0% |
| math 500 | 73.3% | — |
| mmlu pro | 69.0% | 81.9% |
| scicode | 23.3% | 30.7% |
| tau2 | 23.1% | 73.4% |
| terminalbench hard | 4.5% | 23.5% |
Benchmark data from Artificial Analysis.