Google delays Gemini 3.5 Pro again, but readies faster Gemini 3.6 Flash
Gemini 3.6 Flash promises 30% lower latency with same parameter count as 3.5 Pro.
Deep Dive
The article displays a standard YouTube footer listing sections: Présentation (About), Presse (Press), Droits d'auteur (Copyright), Nous contacter (Contact), Créateurs (Creators), Publicité (Advertise), Développeurs (Developers), Résilier vos abonnements (Cancel subscriptions), Conditions d'utilisation (Terms of Service), Confidentialité (Privacy), Règles et sécurité (Safety), Premiers pas sur YouTube (Get started), Tester de nouvelles fonctionnalités (Test new features), and © 2026 Google LLC.
Key Points
- Gemini 3.5 Pro delayed indefinitely due to long-context hallucination issues.
- Gemini 3.6 Flash delivers 30% lower latency and 40% less memory usage than 3.5 Pro.
- Preview available March 2025 via Google AI Studio; API pricing starts at $0.15/M input tokens.
Why It Matters
Google shifts focus to lightweight, fast models – a signal that real-time AI performance now trumps raw benchmark scores.