← All comparisons
Hermes 3 - Llama-3.1 70B vs Qwen3 235B A22B 2507 Instruct
Nous Research vs Alibaba — side-by-side benchmark comparison
| Hermes 3 - Llama-3.1 70B | Qwen3 235B A22B 2507 Instruct | |
|---|---|---|
| Intelligence Index | 10.6 | 25.0 |
| Coding Index | — | 22.1 |
| Math Index | — | 71.7 |
| Output speed (tok/s) | 33.2 | 57.0 |
| Blended price ($/1M) | $0.30 | $0.36 |
| Time to first token (s) | 0.38s | 1.34s |
| aime | 2.3% | 71.7% |
| aime 25 | — | 71.7% |
| artificial analysis coding index | — | 22.10 |
| artificial analysis intelligence index | 10.60 | 25.00 |
| artificial analysis math index | — | 71.70 |
| gpqa | 40.1% | 75.3% |
| hle | 4.1% | 10.6% |
| ifbench | — | 46.1% |
| lcr | — | 31.2% |
| livecodebench | 18.8% | 52.4% |
| math 500 | 53.8% | 98.0% |
| mmlu pro | 57.1% | 82.8% |
| scicode | 23.1% | 36.0% |
| tau2 | — | 33.3% |
| terminalbench hard | — | 15.2% |
Benchmark data from Artificial Analysis.