Gemini 3 Deep Think
Google's enhanced reasoning mode with extended thinking capabilities for complex problems
Gemini 3 Deep Think
Google • November 2025
Training Data
Up to mid 2025
Gemini 3 Deep Think
November 2025
Parameters
~600 billion
Training Method
Deep Reasoning Enhanced Transformer
Context Window
1,000,000 tokens
Knowledge Cutoff
November 2025
Key Features
Deep Reasoning • Extended Thinking • Complex Problem Solving • Research Grade
Capabilities
Reasoning: Maximum
Math: Outstanding (91.9% GPQA Diamond)
Research: Outstanding
What's New in This Version
Extended thinking mode that pushes reasoning performance beyond standard Gemini 3 Pro
Google's enhanced reasoning mode with extended thinking capabilities for complex problems
What's New in This Version
Extended thinking mode that pushes reasoning performance beyond standard Gemini 3 Pro
Technical Specifications
Key Features
Capabilities
Other Google Models
Explore more models from Google
Gemini 4 Argon
Google's first Gemini 4 model and new frontier tier, built for long-horizon software engineering, enterprise knowledge work, and cybersecurity defense — released first to vetted cyber defenders through the Fairwind Program
Gemini 3.8 Flash
Google's most intelligent workhorse model, taking extra reasoning steps and iterative tool calls on hard tasks for gains in software engineering, agentic work, and multi-step reasoning at the same price as Gemini 3.7 Flash
Gemini 3.8 Flash Cyber
A cybersecurity post-training of Gemini 3.8 Flash with more permissive safeguards, offered only to trusted defenders — government authorities, critical infrastructure operators, and software maintainers — through Google's Fairwind Program
