Gemini 4 Argon
AI reference score 8.60 ± 0.52 Last 3 weekly runs · 9 images 1 weekly test leftUpdated every Sunday · 2026-10-25Top tiers of one model
Latest samples
Score trend
high Normal
12 records in the last 30 days
Earlier samples
high
Versions by score
Scores by criterion
Pelican likeness 9.00
Bicycle correctness 8.94
Riding pose 8.67
Composition 9.17
Obvious errors (higher is better) 8.22
Release notes
Original (Chinese)当天向开发者和企业全球开放:Gemini API(gemini-4-argon)、Google AI Studio、Google Antigravity、Gemini for Business,并向 Google AI Pro / Ultra 订阅用户在 Gemini 应用和 AI Mode 中推出。