Introducing Gemini 3.8 Live and 3.8 Live Extended Thinking
Google has released two new model variants: Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking. These models are specifically engineered to facilitate natural, real-time dialogue interactions.
Verified State Diff
Impact & Verification Analysis
Developers utilizing the Gemini API, enterprise users of Google Cloud Vertex AI, and end-users of Google's conversational AI interfaces.
This release signals a strategic move to compete with reasoning-focused models in real-time environments, potentially reducing the latency-accuracy trade-off for complex, multi-turn conversational tasks.
Full Fact Overview
The release of the 3.8 series marks a shift toward specialized live-interaction architectures. 'Extended Thinking' suggests the integration of chain-of-thought or reasoning-heavy inference paths within a low-latency live dialogue framework, likely utilizing a multi-modal streaming architecture to balance conversational fluidity with complex problem-solving capabilities.