Gemini 3.6 Flash (Minimal): benchmark results
Provider: Google. Released 2026-07-21. Access: API.
Unified ELO 1648 ± 1, rank #357 of 3078 rated models, from 20 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| FlavourBench - Substitution | 68.12 | Substitution Family Score (0-100) | 92.3 |
| Chess Puzzles (Epoch AI) | 35 | Accuracy (%) | 83.3 |
| Korean CSAT 2026 (Easy Mode) - English | 98 | Points (out of 100) | 79.6 |
| Korean CSAT 2026 (Easy Mode) - Korean | 98 | Points (out of 100) | 77.2 |
| Epoch AI - GPQA Diamond | 85.86 | Accuracy (%) | 75.5 |
| FlavourBench - Constraint | 51.59 | Constraint Family Score (0-100) | 73.1 |
| Korean CSAT 2026 (Easy Mode) - Total | 423 | Points (out of 450) | 70.7 |
| Epoch AI - Mystery Game Puzzles | 25 | Score | 70.6 |
| Korean CSAT 2026 (Easy Mode) - Society and Culture | 42 | Points (out of 50) | 68 |
| Korean CSAT 2026 (Easy Mode) - Physics I | 38 | Points (out of 50) | 66.7 |
| OTIS Mock AIME 2024-25 | 80 | Accuracy (%) | 63.1 |
| FlavourBench | 61.98 | FlavourBench Score (0-100) | 61.5 |
Interactive version: theaggregate.ai/model?slug=gemini-3-6-flash-minimal · How It Works · Data refreshed daily, snapshot 2026-09-19.