← All comparisons
Claude Sonnet 4.6 (Non-reasoning, High Effort) vs Molmo2-8B
Anthropic vs Allen Institute for AI — side-by-side benchmark comparison
| Claude Sonnet 4.6 (Non-reasoning, High Effort) | Molmo2-8B | |
|---|---|---|
| Intelligence Index | 44.4 | 7.3 |
| Coding Index | 46.4 | 4.4 |
| Math Index | — | — |
| Output speed (tok/s) | 55.2 | 0.0 |
| Blended price ($/1M) | $6.56 | $0.00 |
| Time to first token (s) | 1.07s | 0.00s |
| aime | — | — |
| aime 25 | — | — |
| artificial analysis coding index | 46.40 | 4.40 |
| artificial analysis intelligence index | 44.40 | 7.30 |
| artificial analysis math index | — | — |
| gpqa | 79.9% | 42.5% |
| hle | 13.2% | 4.4% |
| ifbench | 41.2% | 26.9% |
| lcr | 57.7% | 0.0% |
| livecodebench | — | — |
| math 500 | — | — |
| mmlu pro | — | — |
| scicode | 46.9% | 13.3% |
| tau2 | 79.5% | 0.0% |
| terminalbench hard | 46.2% | 0.0% |
Benchmark data from Artificial Analysis.