| Benchmark | Gemini 3 Flash Preview | Nemotron 3 Super 120B A12B |
|---|---|---|
| AA Agentic Index | 49.7 | 8.8 |
| AA Intelligence | 38.7 | 25.7 |
| AA-LCR | 73 | 60.3 |
| AA-Omniscience | 10.1 | -41.5 |
| AIME 2025 | 99.7 | 90.2 |
| aime_2026 | 96.7 | 91.7 |
| apexAgents | 27.7 | 1.8 |
| Artificial Analysis Coding Index | 42.6 | 37.7 |
| critpt | 8.6 | 3.1 |
| gdpval | 35.2 | 9.9 |
| GPQA Diamond | 90.4 | 82.7 |
| HLE | 43.5 | 22.8 |
| hmmt_feb_2026 | 89.4 | 85.2 |
| IFBench | 78 | 72.6 |
| OmniScience Accuracy | 53.4 | 24.3 |
| OmniScience Non-Hallucination | 7.6 | 13 |
| scicode | 50.6 | 42 |
| SWE-bench Multilingual | 72.7 | 45.8 |
| SWE-bench Verified | 78 | 53.7 |
| TauBench V3 - Banking | 20.8 | 10.3 |
| Terminal-Bench 2.0 | 47.6 | 31 |
| Terminal-Bench 2.1 | 58 | 38.6 |
| Terminal-Bench Hard | 38.6 | 28.8 |
| τ²-Bench Telecom (AA run) | 80.4 | 67.8 |
| τ³-Bench | 17.5 | 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.