Google Announces Gemini 3.8 Live
Google has announced Gemini 3.8 Live, a new AI model capable of conversational task execution and persistent background processing even after conversations conc
Image source: https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-3-8-live-gemini-3-8-live-extended-thinking/
On September 16, 2026, Google officially announced Gemini 3.8 Live, a new AI model designed to enhance productivity through conversational task execution and persistent background processing capabilities.
The newly unveiled Gemini 3.8 Live goes beyond traditional response formats by allowing users to coordinate tasks through conversational interaction, maintaining the workflow even after the voice conversation concludes.
Conversational Task Coordination and Background Processing
The most notable feature of Gemini 3.8 Live is its capability for conversational task execution.
- Conversational Task Execution: Users can discuss and coordinate goals with the model using natural language, and Gemini will analyze and execute tasks step-by-step.
- Persistent Workflow: Even after the conversation ends, the model continues to process the assigned tasks in the background without requiring active user monitoring.
While previous conversational AI models were primarily focused on immediate responses to user inputs, this update strengthens the agentic characteristics of the model by enabling it to proactively complete tasks that were negotiated during the conversation.
Evolution for Practical Productivity
Gemini 3.8 Live is expected to provide practical assistance in improving user productivity beyond simple Q&A interactions. The ability to independently process tasks in the background after a conversation ends will likely be beneficial for automating repetitive and time-consuming tasks such as lengthy data processing, complex document summarization, and classification.