← All comparisons
DeepHermes 3 - Llama-3.1 8B Preview (Non-reasoning) vs GPT-4o mini
Nous Research vs OpenAI — side-by-side benchmark comparison
| DeepHermes 3 - Llama-3.1 8B Preview (Non-reasoning) | GPT-4o mini | |
|---|---|---|
| Intelligence Index | 7.6 | 12.6 |
| Coding Index | — | — |
| Math Index | — | 14.7 |
| Output speed (tok/s) | 0.0 | 69.6 |
| Blended price ($/1M) | $0.00 | $0.26 |
| Time to first token (s) | 0.00s | 0.62s |
| aime | 0.0% | 11.7% |
| aime 25 | — | 14.7% |
| artificial analysis coding index | — | — |
| artificial analysis intelligence index | 7.60 | 12.60 |
| artificial analysis math index | — | 14.70 |
| gpqa | 27.0% | 42.6% |
| hle | 4.3% | 4.0% |
| ifbench | — | 31.0% |
| lcr | — | — |
| livecodebench | 8.5% | 23.4% |
| math 500 | 21.8% | 78.9% |
| mmlu pro | 36.5% | 64.8% |
| scicode | 9.1% | 22.9% |
| tau2 | — | — |
| terminalbench hard | — | — |
Benchmark data from Artificial Analysis.