| Benchmark | Gemini 2.5 Flash | Nemotron 3 Super 120B A12B |
|---|---|---|
| AA Agentic Index | 18.8 | 8.8 |
| AA Intelligence | 17.7 | 25.7 |
| AA-LCR | 65.7 | 60.3 |
| AA-Omniscience | -29.8 | -41.5 |
| AIME 2025 | 72 | 90.2 |
| Artificial Analysis Coding Index | 22.2 | 37.7 |
| critpt | 1.4 | 3.1 |
| gdpval | 11.9 | 9.9 |
| GPQA Diamond | 82.8 | 82.7 |
| HLE | 12.1 | 22.8 |
| IFBench | 50.3 | 72.6 |
| LiveCodeBench | 76.2 | 82.1 |
| OmniScience Accuracy | 26.1 | 24.3 |
| OmniScience Non-Hallucination | 24.7 | 13 |
| scicode | 35.9 | 42 |
| SWE-bench Verified | 60.4 | 53.7 |
| Terminal-Bench Hard | 13.6 | 28.8 |
| τ²-Bench Telecom (AA run) | 31.6 | 67.8 |
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.