← All comparisons
NVIDIA Nemotron Nano 9B V2 (Reasoning) vs DeepSeek R1 Distill Llama 8B
NVIDIA vs DeepSeek — side-by-side benchmark comparison
| NVIDIA Nemotron Nano 9B V2 (Reasoning) | DeepSeek R1 Distill Llama 8B | |
|---|---|---|
| Intelligence Index | 14.8 | 12.1 |
| Coding Index | 8.3 | — |
| Math Index | 69.7 | 41.3 |
| Output speed (tok/s) | 123.4 | 0.0 |
| Blended price ($/1M) | $0.07 | $0.00 |
| Time to first token (s) | 0.24s | 0.00s |
| aime | — | 33.3% |
| aime 25 | 69.7% | 41.3% |
| artificial analysis coding index | 8.30 | — |
| artificial analysis intelligence index | 14.80 | 12.10 |
| artificial analysis math index | 69.70 | 41.30 |
| gpqa | 57.0% | 30.2% |
| hle | 4.6% | 4.2% |
| ifbench | 27.6% | 17.6% |
| lcr | 21.0% | 0.0% |
| livecodebench | 72.4% | 23.3% |
| math 500 | — | 85.3% |
| mmlu pro | 74.2% | 54.3% |
| scicode | 22.0% | 11.9% |
| tau2 | 21.9% | — |
| terminalbench hard | 1.5% | — |
Benchmark data from Artificial Analysis.