Gemini 3.5 Flash Lite (High) — benchmark results
Provider: Google. Released 2026-07-21. Access: API.
Unified ELO 1755 ± 35, rank #227 of 1806 rated models, from 9 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Chess Puzzles (Epoch AI) | 22 | Accuracy (%) | 56.3 |
| ARC-AGI-2 | 10.28 | Accuracy (%) | 55.2 |
| OTIS Mock AIME 2024-25 | 71.11 | Accuracy (%) | 52.3 |
| LLM2014 Logic 2026-07 | 22.82 | Median Score | 51.2 |
| ARC-AGI-1 | 53.5 | Accuracy (%) | 46.8 |
| WeirdML | 39 | Average Score | 33.3 |
| Riemann-bench | 11.2 | Score (%) | 22.9 |
| GDP.pdf | 10 | Strict Pass Rate (%) | 17.3 |
| DeepsecBench | 1.07 | Recall-weighted F2 score (%) | 0 |
Interactive version: theaggregate.ai/model?slug=gemini-3-5-flash-lite-high · How It Works · Data refreshed daily, snapshot 2026-08-07.