Gemini 3.5 Flash (Medium) — benchmark results

Gemini 3.5 Flash evaluated at the medium reasoning-effort setting. Provider: Google. Released 2026-05-19. Access: API.

Unified ELO 1942 ± 23, rank #44 of 1776 rated models, from 49 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
AA MMMU-Pro83.87Accuracy (%)99.1
AA Omniscience - Law57.5Accuracy (%)99.1
AA Omniscience21.65Score98.2
AA Omniscience - Humanities & Social Sciences53Accuracy (%)97.4
AA Omniscience - Software Engineering (SWE) - PHP82Accuracy (%)97.2
AA GPQA Diamond92.12Accuracy (%)97.1
Chess Bench LLM1957Lichess Rating97
AA-LCR71Score (self-reported)96.7
AA Humanity's Last Exam39.85Accuracy (%)96.5
AA Omniscience - Business42.7Accuracy (%)96.4
AA-Omniscience Accuracy51.12Accuracy (%)96.2
AA Omniscience - Science, Engineering & Mathematics48.3Accuracy (%)96

Interactive version: theaggregate.ai/model?slug=gemini-3-5-flash-medium · How the rankings work · Data refreshed daily, snapshot 2026-07-22.