| Benchmark | Muse Spark 1.1 | Qwen2.5 Instruct 72B |
|---|---|---|
| AA Intelligence | 53.2 | 10 |
| AA-LCR | 81.3 | 20.3 |
| AA-Omniscience | 28.1 | -52.2 |
| Artificial Analysis Coding Index | 71.3 | 11.9 |
| critpt | 15.1 | 0 |
| Fortress | 12.4 | 56.4 |
| GPQA Diamond | 89.8 | 49.1 |
| HLE | 62.1 | 4.2 |
| OmniScience Accuracy | 52 | 17.6 |
| OmniScience Non-Hallucination | 50 | 15.3 |
| scicode | 58.2 | 26.7 |
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.