Gemini 3.7 Flash: benchmark results
Google's Gemini 3.7 Flash (August 13, 2026), an algorithmic refinement of 3.6 Flash tuned for software engineering, agent workflows and multi-step execution, with the same 1,048,576-token context window. Provider: Google. Released 2026-08-13. Access: API.
Aggregate Rating 1872 ± 5, rank #23 of 1656 rated models, from 679 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA-AnalystAgent | 60 | Score (%) | 100 |
| AA-AnalystAgent - Pass@1 | 70.5 | Pass@1 (%) | 100 |
| AI for Education Visual Reasoning | 89.9 | Accuracy (%) | 100 |
| AI for Education Visual Reasoning - match (figure) | 87.4 | Accuracy (%) | 100 |
| AI for Education Visual Reasoning - pattern completion (linear) | 95.7 | Accuracy (%) | 100 |
| AI for Education Visual Reasoning - reasoning by analogy | 94.4 | Accuracy (%) | 100 |
| CESBench - Side-Channel | 85.5 | Composite score (%; mean over the four task types of the ite | 100 |
| CraftBench-UE (C++, Aura MCP) | 93.94 | Task completion (%; C++ tasks completed within 40 minutes ov | 100 |
| DSV-Mem | 39.6 | Accuracy (%), 1,000 stateful visual-memory questions, mean o | 100 |
| DroneCATS (Approaching) | 65 | Success Rate (%) | 100 |
| EnterpriseOps-Gym-AA - ITSM | 50.28 | Task Success (%) | 100 |
| JudgeSense - Relevance | 0.98 | Judge Sensitivity Score: share of items on which the judge r | 100 |
Interactive version: theaggregate.ai/model?slug=gemini-3-7-flash · How It Works · Data refreshed daily, snapshot 2026-10-20.