← All comparisons
GPT-5.4 nano (medium) vs Hermes 4 - Llama-3.1 405B (Non-reasoning)
OpenAI vs Nous Research — side-by-side benchmark comparison
| GPT-5.4 nano (medium) | Hermes 4 - Llama-3.1 405B (Non-reasoning) | |
|---|---|---|
| Intelligence Index | 38.1 | 17.6 |
| Coding Index | 35.0 | 18.1 |
| Math Index | — | 15.3 |
| Output speed (tok/s) | 152.6 | 40.8 |
| Blended price ($/1M) | $0.46 | $1.50 |
| Time to first token (s) | 2.09s | 0.73s |
| aime | — | — |
| aime 25 | — | 15.3% |
| artificial analysis coding index | 35.00 | 18.10 |
| artificial analysis intelligence index | 38.10 | 17.60 |
| artificial analysis math index | — | 15.30 |
| gpqa | 76.1% | 53.6% |
| hle | 14.7% | 4.2% |
| ifbench | 64.4% | 34.8% |
| lcr | 57.3% | 20.0% |
| livecodebench | — | 54.6% |
| math 500 | — | — |
| mmlu pro | — | 72.9% |
| scicode | 38.4% | 34.6% |
| tau2 | 52.6% | 26.6% |
| terminalbench hard | 33.3% | 9.8% |
Benchmark data from Artificial Analysis.