← All comparisons
GPT-5.4 mini (Non-Reasoning) vs Hermes 3 - Llama-3.1 70B
OpenAI vs Nous Research — side-by-side benchmark comparison
| GPT-5.4 mini (Non-Reasoning) | Hermes 3 - Llama-3.1 70B | |
|---|---|---|
| Intelligence Index | 23.3 | 10.6 |
| Coding Index | 25.3 | — |
| Math Index | — | — |
| Output speed (tok/s) | 166.4 | 33.2 |
| Blended price ($/1M) | $1.69 | $0.30 |
| Time to first token (s) | 0.48s | 0.38s |
| aime | — | 2.3% |
| aime 25 | — | — |
| artificial analysis coding index | 25.30 | — |
| artificial analysis intelligence index | 23.30 | 10.60 |
| artificial analysis math index | — | — |
| gpqa | 60.6% | 40.1% |
| hle | 5.7% | 4.1% |
| ifbench | 38.8% | — |
| lcr | 30.3% | — |
| livecodebench | — | 18.8% |
| math 500 | — | 53.8% |
| mmlu pro | — | 57.1% |
| scicode | 39.6% | 23.1% |
| tau2 | 23.4% | — |
| terminalbench hard | 18.2% | — |
Benchmark data from Artificial Analysis.