| Benchmark | Gemini 3 Flash Preview | Qwen3.6 27B |
|---|---|---|
| AA Agentic Index | 49.7 | 27.5 |
| AA Intelligence | 38.7 | 37.7 |
| AA-LCR | 73 | 73.3 |
| AA-Omniscience | 10.1 | -20 |
| aime_2026 | 96.7 | 94.1 |
| Artificial Analysis Coding Index | 42.6 | 53.7 |
| charxiv_rq | 80.3 | 78.4 |
| critpt | 8.6 | 1.1 |
| gdpval | 35.2 | 32 |
| GPQA Diamond | 90.4 | 87.8 |
| HLE | 43.5 | 24 |
| hmmt_feb_2025 | 97.5 | 93.8 |
| hmmt_feb_2026 | 89.4 | 84.3 |
| hmmt_nov_2025 | 93.3 | 90.7 |
| IFBench | 78 | 69.1 |
| MMMU-Pro | 81.2 | 75.8 |
| OmniDocBench 1.5 | 12.1 | 89.4 |
| OmniScience Accuracy | 53.4 | 19.6 |
| OmniScience Non-Hallucination | 7.6 | 50.7 |
| OSWorld-Verified | 65.1 | 63.9 |
| scicode | 50.6 | 39.8 |
| simpleqa | 68.7 | 95.7 |
| SWE-bench Multilingual | 72.7 | 71.3 |
| SWE-bench Pro | 49.6 | 53.5 |
| SWE-bench Verified | 78 | 77.2 |
| TauBench V3 - Banking | 20.8 | 16.7 |
| Terminal-Bench 2.0 | 47.6 | 59.3 |
| Terminal-Bench 2.1 | 58 | 60.7 |
| Terminal-Bench Hard | 38.6 | 34.8 |
| VideoMMMU | 86.9 | 84.4 |
| τ²-Bench Telecom (AA run) | 80.4 | 94.2 |
| τ³-Bench | 17.5 | 15.3 |
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.