| Benchmark | Claude 3.7 Sonnet | Muse Spark 1.1 |
|---|---|---|
| AA Agentic Index | 37 | 39.7 |
| AA Intelligence | 27.6 | 53.2 |
| AA-LCR | 62.3 | 81.3 |
| AA-Omniscience | -0.7 | 28.1 |
| Artificial Analysis Coding Index | 36.4 | 71.3 |
| critpt | 0.9 | 15.1 |
| Fortress | 38 | 12.4 |
| gdpval | 27.4 | 43.7 |
| GPQA Diamond | 84.8 | 89.8 |
| HLE | 9.7 | 62.1 |
| multichallenge | 51.6 | 75.3 |
| MultiNRC | 27.8 | 65.6 |
| OmniScience Accuracy | 28.2 | 52 |
| OmniScience Non-Hallucination | 60 | 50 |
| scicode | 40.3 | 58.2 |
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.