← All comparisons
Llama 3.1 Nemotron Instruct 70B vs Molmo2-8B
NVIDIA vs Allen Institute for AI — side-by-side benchmark comparison
| Llama 3.1 Nemotron Instruct 70B | Molmo2-8B | |
|---|---|---|
| Intelligence Index | 13.4 | 7.3 |
| Coding Index | 10.8 | 4.4 |
| Math Index | 11.0 | — |
| Output speed (tok/s) | 292.9 | 0.0 |
| Blended price ($/1M) | $1.20 | $0.00 |
| Time to first token (s) | 0.24s | 0.00s |
| aime | 24.7% | — |
| aime 25 | 11.0% | — |
| artificial analysis coding index | 10.80 | 4.40 |
| artificial analysis intelligence index | 13.40 | 7.30 |
| artificial analysis math index | 11.00 | — |
| gpqa | 46.5% | 42.5% |
| hle | 4.6% | 4.4% |
| ifbench | 30.7% | 26.9% |
| lcr | 7.0% | 0.0% |
| livecodebench | 16.9% | — |
| math 500 | 73.3% | — |
| mmlu pro | 69.0% | — |
| scicode | 23.3% | 13.3% |
| tau2 | 23.1% | 0.0% |
| terminalbench hard | 4.5% | 0.0% |
Benchmark data from Artificial Analysis.