Google Gemini Live Avatar is now available to businesses with a Gemini Enterprise subscription, giving conversational AI agents a visible on-screen presence. The rollout follows the recent announcement of Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking.
The feature combines an interactive visual experience with a conversational AI agent, allowing customers to complete tasks such as filing an insurance claim through a video chat. Businesses and Google Enterprise users can select from a range of avatars to put a face to an AI agent’s voice, whether cartoon-style or more realistic in appearance.
What Live Avatar Can Do
Live Avatar creates live video avatars with lip synching for conversational video, and can process live audio, screen sharing and video feeds at the same time. It can also run tools and make API calls while continuing to speak with customers, so an interaction can move forward without interruption.
The tool uses native speech-to-speech to deliver more fluid dialogue. According to Google, this is designed to allow more natural interruption recovery without dropping conversation context or backend transactions, meaning a customer can talk over the avatar without the exchange losing its thread.
Safeguards and Availability
Google has put several safeguards in place to reduce the risk of misuse. Businesses can choose from curated, prebuilt avatars, while anyone wishing to create a custom avatar must complete a verification process. All audio and video generated through the feature are watermarked with SynthID, which is intended to keep AI-generated content verifiable.
Some questions remain open. Google did not immediately confirm when Live Avatar will be available to the general public, or how it plans to protect the information an AI agent receives during video calls. In the insurance claim example, that could include a customer’s name, address and live video footage captured while completing the process.
The feature is currently offered to organisations with a Gemini Enterprise subscription.
