← All comparisons
Hermes 4 - Llama-3.1 70B (Non-reasoning) vs DeepSeek R1 Distill Qwen 1.5B
Nous Research vs DeepSeek — side-by-side benchmark comparison
| Hermes 4 - Llama-3.1 70B (Non-reasoning) | DeepSeek R1 Distill Qwen 1.5B | |
|---|---|---|
| Intelligence Index | 12.6 | 9.1 |
| Coding Index | 9.2 | — |
| Math Index | 11.3 | 22.0 |
| Output speed (tok/s) | 94.3 | 0.0 |
| Blended price ($/1M) | $0.20 | $0.00 |
| Time to first token (s) | 0.61s | 0.00s |
| aime | — | 17.7% |
| aime 25 | 11.3% | 22.0% |
| artificial analysis coding index | 9.20 | — |
| artificial analysis intelligence index | 12.60 | 9.10 |
| artificial analysis math index | 11.30 | 22.00 |
| gpqa | 49.1% | 9.8% |
| hle | 3.6% | 3.3% |
| ifbench | 29.0% | 13.2% |
| lcr | 2.0% | 0.3% |
| livecodebench | 26.9% | 7.0% |
| math 500 | — | 68.7% |
| mmlu pro | 66.4% | 26.9% |
| scicode | 27.7% | 6.6% |
| tau2 | 21.6% | — |
| terminalbench hard | 0.0% | — |
Benchmark data from Artificial Analysis.