← All comparisons
GPT-5.4 nano (Non-Reasoning) vs Llama 3.1 Nemotron Ultra 253B v1 (Reasoning)
OpenAI vs NVIDIA — side-by-side benchmark comparison
| GPT-5.4 nano (Non-Reasoning) | Llama 3.1 Nemotron Ultra 253B v1 (Reasoning) | |
|---|---|---|
| Intelligence Index | 24.4 | 15.0 |
| Coding Index | 27.9 | 13.1 |
| Math Index | — | 63.7 |
| Output speed (tok/s) | 157.4 | 52.3 |
| Blended price ($/1M) | $0.46 | $0.90 |
| Time to first token (s) | 0.54s | 0.76s |
| aime | — | 74.7% |
| aime 25 | — | 63.7% |
| artificial analysis coding index | 27.90 | 13.10 |
| artificial analysis intelligence index | 24.40 | 15.00 |
| artificial analysis math index | — | 63.70 |
| gpqa | 55.8% | 72.8% |
| hle | 4.2% | 8.1% |
| ifbench | 32.7% | 38.2% |
| lcr | 24.7% | 7.3% |
| livecodebench | — | 64.1% |
| math 500 | — | 95.2% |
| mmlu pro | — | 82.5% |
| scicode | 35.2% | 34.7% |
| tau2 | 34.8% | 11.4% |
| terminalbench hard | 24.2% | 2.3% |
Benchmark data from Artificial Analysis.