| Benchmark | Nemotron 3 Super 120B A12B | o1 |
|---|---|---|
| AA Agentic Index | 8.8 | 31.1 |
| AA Intelligence | 25.7 | 24 |
| AA-LCR | 60.3 | 63.3 |
| AA-Omniscience | -41.5 | -10.6 |
| AIME 2025 | 90.2 | 81.7 |
| Artificial Analysis Coding Index | 37.7 | 39.7 |
| critpt | 3.1 | 0.3 |
| gdpval | 9.9 | 11.5 |
| GPQA Diamond | 82.7 | 75.7 |
| GSM8K | 90.7 | 97.1 |
| HLE | 22.8 | 7.7 |
| humaneval | 80.5 | 88.1 |
| IFBench | 72.6 | 70.3 |
| mmlu | 86 | 91.8 |
| OmniScience Accuracy | 24.3 | 34.7 |
| OmniScience Non-Hallucination | 13 | 30.7 |
| scicode | 42 | 35.8 |
| SWE-bench Verified | 53.7 | 48.9 |
| Terminal-Bench Hard | 28.8 | 12.9 |
| τ²-Bench Telecom (AA run) | 67.8 | 62.6 |
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.