← All comparisons
NVIDIA Nemotron 3 Nano 30B A3B (Reasoning) vs Llama 3.3 Nemotron Super 49B v1 (Reasoning)
NVIDIA vs NVIDIA — side-by-side benchmark comparison
| NVIDIA Nemotron 3 Nano 30B A3B (Reasoning) | Llama 3.3 Nemotron Super 49B v1 (Reasoning) | |
|---|---|---|
| Intelligence Index | 24.3 | 18.5 |
| Coding Index | 19.0 | 9.4 |
| Math Index | 91.0 | 54.7 |
| Output speed (tok/s) | 122.8 | 0.0 |
| Blended price ($/1M) | $0.10 | $0.00 |
| Time to first token (s) | 0.83s | 0.00s |
| aime | — | 58.3% |
| aime 25 | 91.0% | 54.7% |
| artificial analysis coding index | 19.00 | 9.40 |
| artificial analysis intelligence index | 24.30 | 18.50 |
| artificial analysis math index | 91.00 | 54.70 |
| gpqa | 75.7% | 64.3% |
| hle | 10.2% | 6.5% |
| ifbench | 71.1% | 38.1% |
| lcr | 33.7% | 17.0% |
| livecodebench | 74.1% | 27.7% |
| math 500 | — | 95.9% |
| mmlu pro | 79.4% | 78.5% |
| scicode | 29.6% | 28.2% |
| tau2 | 40.9% | 26.9% |
| terminalbench hard | 13.6% | 0.0% |
Benchmark data from Artificial Analysis.