| Benchmark | Claude Sonnet 5 | Llama 3.1 Instruct 405B |
|---|---|---|
| AA Agentic Index | 49.7 | 6.3 |
| AA Intelligence | 55.3 | 8.5 |
| AA-LCR | 77 | 24.7 |
| AA-Omniscience | 16.4 | -17.1 |
| Artificial Analysis Coding Index | 71.5 | 14.5 |
| critpt | 16.9 | 0 |
| gdpval | 54.8 | 0 |
| GPQA Diamond | 91.1 | 51.5 |
| HLE | 57.4 | 4.2 |
| OmniScience Accuracy | 40 | 23.2 |
| OmniScience Non-Hallucination | 60.6 | 49 |
| scicode | 53.6 | 29.9 |
| simplebench | 60.6 | 23 |
| SWE-bench Pro | 63.2 | 11.2 |
| SWE-bench Verified | 85.2 | 24.5 |
Best tracked score per model per benchmark (default configuration; source-attributed). ↓ marks lower-is-better metrics. Open either model for its full surface, provenance and pricing.