Gemini 2.5 Flash (Preview 05-20) (Thinking): benchmark results

Provider: Google. Released 2025-05-20. Access: API.

Unified ELO 1647 ± 1, rank #360 of 3078 rated models, from 9 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
BenchTable - Utility82.2Weighted Score (%)97.4
DuckDB-NSQL73.3Execution Accuracy (%)92.2
BenchTable - Tech79.6Weighted Score (%)85
BenchTable - Reasoning66.9Weighted Score (%)82.3
BenchTable66.6Total Score (%)81.1
SpeechMap Compliance71.8% Requests Completed66.1
BenchTable - STEM56.2Weighted Score (%)56.1
JEE/NEET LLM Benchmark - NEET 202595.14Exam Score (%)50
JEE/NEET LLM Benchmark - JEE Advanced 202581.94Exam Score (%)25

Interactive version: theaggregate.ai/model?slug=gemini-2-5-flash-preview-05-20-thinking · How It Works · Data refreshed daily, snapshot 2026-09-19.