Gemini 3.1 Flash Lite (Low): benchmark results
Provider: Google. Released 2026-05-07. Access: API.
Unified ELO 1582 ± 1, rank #733 of 3081 rated models, from 11 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Chess Puzzles (Epoch AI) | 25 | Accuracy (%) | 71.7 |
| SLR-Bench - Easy | 76 | Accuracy (%) | 51.8 |
| SLR-Bench - Basic | 98 | Accuracy (%) | 50 |
| Epoch AI - GPQA Diamond | 74.24 | Accuracy (%) | 49.7 |
| SLR-Bench - Medium | 13 | Accuracy (%) | 43.6 |
| SLR-Bench - Hard | 3 | Accuracy (%) | 42.7 |
| FrontierMath - Tiers 1-3 (v2) | 22.46 | Accuracy (%, 285 private v2 problems) | 36.9 |
| ObviousBench | 83.33 | Answer pass³ (%) | 36.9 |
| OTIS Mock AIME 2024-25 | 44.44 | Accuracy (%) | 32.8 |
| GIM | -0.97 | IRT ability (theta) | 11.1 |
| DeepResearchBench | 37.3 | Average Score | 10 |
Interactive version: theaggregate.ai/model?slug=gemini-3-1-flash-lite-low · How It Works · Data refreshed daily, snapshot 2026-09-21.