| Benchmark | Gemini 3 Flash Preview | Qwen3.7 Max |
|---|---|---|
| AA Agentic Index | 49.7 | 30.9 |
| AA Intelligence | 38.7 | 46.7 |
| AA-LCR | 73 | 74.7 |
| AA-Omniscience | 10.1 | 13.5 |
| aime_2026 | 96.7 | 97 |
| arena_elo | 1473 | 1474 |
| arena_text_factuality | 1452 | 1479 |
| Artificial Analysis Coding Index | 42.6 | 66 |
| coding_arena_elo | 1387 | 1517 |
| critpt | 8.6 | 13.4 |
| Finance Agent v2 | 42.6 | 48.4 |
| gdpval | 35.2 | 38.5 |
| GPQA Diamond | 90.4 | 92.4 |
| HLE | 43.5 | 41.4 |
| hmmt_feb_2026 | 89.4 | 97.1 |
| hmmt_nov_2025 | 93.3 | 95 |
| IFBench | 78 | 80.5 |
| livebench | 72.4 | 74.3 |
| MCP Atlas | 62 | 76.4 |
| mmmlu | 91.8 | 90.3 |
| OmniScience Accuracy | 53.4 | 31.1 |
| OmniScience Non-Hallucination | 7.6 | 74.4 |
| scicode | 50.6 | 53.5 |
| simplebench | 61.1 | 70.4 |
| simpleqa_verified | 67.4 | 58.5 |
| SWE-bench Multilingual | 72.7 | 78.3 |
| SWE-bench Pro | 49.6 | 60.6 |
| SWE-bench Verified | 78 | 80.4 |
| TauBench V3 - Banking | 20.8 | 11.8 |
| Terminal-Bench 2.0 | 47.6 | 69.7 |
| Terminal-Bench 2.1 | 58 | 75 |
| Terminal-Bench Hard | 38.6 | 50.8 |
| τ²-Bench Telecom (AA run) | 80.4 | 94.7 |
| τ³-Bench | 17.5 | 10.9 |
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.