| Benchmark | Gemini 2.5 Flash | Gemma 4 E2B |
|---|---|---|
| AA Agentic Index | 18.8 | 7.4 |
| AA Intelligence | 17.7 | 10 |
| AA-LCR | 65.7 | 17 |
| AA-Omniscience | -29.8 | -23.6 |
| Artificial Analysis Coding Index | 22.2 | 8.3 |
| critpt | 1.4 | 0 |
| gdpval | 11.9 | 0 |
| GPQA Diamond | 82.8 | 43.4 |
| HLE | 12.1 | 4.8 |
| IFBench | 50.3 | 38 |
| MMMU-Pro | 69.1 | 44.6 |
| OmniScience Accuracy | 26.1 | 8.2 |
| OmniScience Non-Hallucination | 24.7 | 67.6 |
| scicode | 35.9 | 20.9 |
| Terminal-Bench Hard | 13.6 | 3 |
| τ²-Bench Telecom (AA run) | 31.6 | 22.2 |
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.