← All comparisons
Hermes 4 - Llama-3.1 70B (Non-reasoning) vs HyperCLOVA X SEED Think (32B)
Nous Research vs Naver — side-by-side benchmark comparison
| Hermes 4 - Llama-3.1 70B (Non-reasoning) | HyperCLOVA X SEED Think (32B) | |
|---|---|---|
| Intelligence Index | 12.6 | 23.7 |
| Coding Index | 9.2 | 17.5 |
| Math Index | 11.3 | 59.0 |
| 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 | — | — |
| aime 25 | 11.3% | 59.0% |
| artificial analysis coding index | 9.20 | 17.50 |
| artificial analysis intelligence index | 12.60 | 23.70 |
| artificial analysis math index | 11.30 | 59.00 |
| gpqa | 49.1% | 61.5% |
| hle | 3.6% | 5.5% |
| ifbench | 29.0% | 37.9% |
| lcr | 2.0% | 11.7% |
| livecodebench | 26.9% | 62.9% |
| math 500 | — | — |
| mmlu pro | 66.4% | 78.5% |
| scicode | 27.7% | 28.4% |
| tau2 | 21.6% | 87.4% |
| terminalbench hard | 0.0% | 12.1% |
Benchmark data from Artificial Analysis.