| Benchmark | Gemini 3.5 Flash | Qwen3.7 Plus Preview |
|---|---|---|
| AA Agentic Index | 70.4 | 20.8 |
| AA Intelligence | 52 | 39.4 |
| AA-LCR | 81 | 69 |
| AA-Omniscience | 21.2 | 2.4 |
| apexAgents | 47.1 | 22.4 |
| arena_elo | 1477 | 1458 |
| arena_text_factuality | 1466 | 1454 |
| Artificial Analysis Coding Index | 70.1 | 55.9 |
| charxiv_rq | 84.2 | 85.9 |
| critpt | 13.1 | 9.1 |
| DeepSWE 1.1 | 37 | 14.2 |
| Finance Agent v2 | 57.9 | 38.2 |
| gdpval | 57.8 | 22.2 |
| GPQA Diamond | 92.2 | 90.3 |
| HLE | 42.7 | 35.6 |
| hmmt_feb_2026 | 95.5 | 92.9 |
| IFBench | 76.3 | 79.1 |
| MCP Atlas | 83.6 | 73.2 |
| MMMU-Pro | 84.3 | 80.5 |
| OmniScience Accuracy | 51.4 | 22.5 |
| OmniScience Non-Hallucination | 38.2 | 74.5 |
| OSWorld-Verified | 78.4 | 73.3 |
| scicode | 53.1 | 51.3 |
| SWE-bench Pro | 55.1 | 57.6 |
| TauBench V3 - Banking | 32.2 | 17.9 |
| Terminal-Bench 2.0 | 76.2 | 70.3 |
| Terminal-Bench 2.1 | 78.7 | 61 |
| Terminal-Bench Hard | 46.2 | 47 |
| τ²-Bench Telecom (AA run) | 95.6 | 93 |
| τ³-Bench | 25.4 | 17.9 |
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.