| Benchmark | Muse Spark 1.1 | Nemotron 3 Super 120B A12B |
|---|---|---|
| AA Agentic Index | 39.7 | 8.8 |
| AA Intelligence | 53.2 | 25.7 |
| AA-LCR | 81.3 | 60.3 |
| AA-Omniscience | 28.1 | -41.5 |
| Artificial Analysis Coding Index | 71.3 | 37.7 |
| critpt | 15.1 | 3.1 |
| gdpval | 43.7 | 9.9 |
| GPQA Diamond | 89.8 | 82.7 |
| HLE | 62.1 | 22.8 |
| multichallenge | 75.3 | 55.2 |
| OmniScience Accuracy | 52 | 24.3 |
| OmniScience Non-Hallucination | 50 | 13 |
| scicode | 58.2 | 42 |
| TauBench V3 - Banking | 31.8 | 10.3 |
| Terminal-Bench 2.1 | 80 | 38.6 |
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.