Gemini 3 Pro — benchmark results

Google's Pro-tier Gemini 3 model for advanced reasoning and general tasks. Provider: Google. Released 2025-11-18. Access: API.

Unified ELO 1847 ± 13, rank #105 of 1776 rated models, from 223 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
BALROG MiniHack (LLM)40Progress (%)100
BALROG NetHack (LLM)6.8Progress (%)100
Diplomacy: Overall Performance60.3Score100
EmoBench-M70.5Average Score100
JMMMU-Pro - Culture Agnostic80.42Accuracy (%)100
JMMMU-Pro - Culture Specific95Accuracy (%)100
JMMMU-Pro - Japanese Art91.33Accuracy (%)100
JMMMU-Pro - Japanese Heritage96.67Accuracy (%)100
JMMMU-Pro - Japanese History95.33Accuracy (%)100
JMMMU-Pro - Overall87.05Accuracy (%)100
JMMMU-Pro - World History96.67Accuracy (%)100
LLM Stats (VideoMMMU)87.6Score (%)100

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