| Benchmark | Gemini 3.5 Flash | Nemotron 3 Super 120B A12B |
|---|---|---|
| AA Agentic Index | 70.4 | 8.8 |
| AA Intelligence | 52 | 25.7 |
| AA-LCR | 81 | 60.3 |
| AA-Omniscience | 21.2 | -41.5 |
| aime_2026 | 95 | 91.7 |
| apexAgents | 47.1 | 1.8 |
| Artificial Analysis Coding Index | 70.1 | 37.7 |
| critpt | 13.1 | 3.1 |
| gdpval | 57.8 | 9.9 |
| GPQA Diamond | 92.2 | 82.7 |
| HLE | 42.7 | 22.8 |
| hmmt_feb_2026 | 95.5 | 85.2 |
| IFBench | 76.3 | 72.6 |
| itbenchSre | 40.3 | 1.1 |
| OmniScience Accuracy | 51.4 | 24.3 |
| OmniScience Non-Hallucination | 38.2 | 13 |
| scicode | 53.1 | 42 |
| TauBench V3 - Banking | 32.2 | 10.3 |
| Terminal-Bench 2.0 | 76.2 | 31 |
| Terminal-Bench 2.1 | 78.7 | 38.6 |
| Terminal-Bench Hard | 46.2 | 28.8 |
| τ²-Bench Telecom (AA run) | 95.6 | 67.8 |
| τ³-Bench | 25.4 | 10.1 |
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.