Gemini 3.1 Flash Lite (Minimal): benchmark results
Provider: Google. Released 2026-05-07. Access: API.
Unified ELO 1641 ± 1, rank #389 of 3078 rated models, from 11 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| SteerBench-Work | 92.8 | Mean Trial Accuracy (%) | 100 |
| SteerBench-Work - Pass^5 | 91.5 | Scenarios Correct in All 5 Trials (%) | 100 |
| Korean CSAT 2026 (Easy Mode) - English | 97 | Points (out of 100) | 66 |
| Korean CSAT 2026 (Easy Mode) - Korean | 90.5 | Points (out of 100) | 46.9 |
| Korean CSAT 2026 (Easy Mode) - Physics I | 28 | Points (out of 50) | 44.6 |
| FrontierMath - Tiers 1-3 (v2) | 21.4 | Accuracy (%, 285 private v2 problems) | 35.5 |
| Korean CSAT 2026 (Easy Mode) - Total | 373 | Points (out of 450) | 35 |
| Korean CSAT 2026 (Easy Mode) - Chemistry I | 30 | Points (out of 50) | 22.1 |
| Korean CSAT 2026 (Easy Mode) - Life Science I | 28 | Points (out of 50) | 18 |
| Korean CSAT 2026 (Easy Mode) - Mathematics | 80 | Points (out of 100) | 16.3 |
| Korean CSAT 2026 (Easy Mode) - Society and Culture | 25 | Points (out of 50) | 15 |
Interactive version: theaggregate.ai/model?slug=gemini-3-1-flash-lite-minimal · How It Works · Data refreshed daily, snapshot 2026-09-19.