Gemini 3.5 Flash — benchmark results
Google's speed-oriented Gemini 3.5 Flash model, balancing capability with lower latency. Provider: Google. Released 2026-05-19. Access: API.
Unified ELO 1887 ± 14, rank #71 of 1776 rated models, from 204 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AI for Education Visual Reasoning | 86 | Accuracy (%) | 100 |
| AI for Education Visual Reasoning - match (figure) | 85.2 | Accuracy (%) | 100 |
| AI for Education Visual Reasoning - reasoning by analogy | 88.8 | Accuracy (%) | 100 |
| APEX-Agents | 66.1 | Mean Score (ReAct) (self-reported) | 100 |
| APEX-Agents-AA | 47.1 | Pass@1 (self-reported) | 100 |
| Arabic Broad Leaderboard | 9.25 | Average Score (0-10) | 100 |
| Benchmarks.bio - BioSecBench-Refusal | 51.5 | Pass Rate (%) | 100 |
| CAIS Vision Capabilities Index | 65.7 | Vision Capabilities Index (self-reported) | 100 |
| Finance Agent v2 | 57.9 | Score (self-reported) | 100 |
| HieroglyphBench | 52.5 | Sign accuracy (%) | 100 |
| MCP Atlas | 83.6 | Score (self-reported) | 100 |
| Vals AI Finance Agent v2 | 57.86 | Accuracy (%) | 100 |
Interactive version: theaggregate.ai/model?slug=gemini-3-5-flash · How the rankings work · Data refreshed daily, snapshot 2026-07-22.