Power your agents: Gemini 3.8 Live with Live Avatar is now generally available

Google Cloud announced the general availability of Gemini 3.8 Live with Live Avatar in Gemini Enterprise, expanding its enterprise voice AI offerings. The technology, first previewed at Google Cloud Next 2026, introduces an interactive visual presence to conversational video agents, supporting deployment across web, mobile, and interactive kiosks. The feature is designed to enhance interaction quality beyond basic speed and cost metrics, providing a native speech-to-speech foundation for fluid dialogue. Enterprises can now access the model with US and EU endpoints, provisioned throughput, and strict data governance standards.

Live Avatar enables customers to deploy pre-built avatars or create custom ones through a controlled enterprise allowlisting and verification process. All generated audio and video streams include imperceptible SynthID watermarks to ensure transparency and verifiability of AI-generated content. The feature allows users to build custom avatars by adding system instructions and uploading a reference photo and audio sample. A demonstration highlights a use case where an insurance claims agent processes damage reports in real time, filling claim notebooks while background agents apply policy rules and build adjuster packets.

Developers can leverage the Agent Development Kit (ADK) to define agents, manage runners, session memory, and stream real-time audio directly to the Gemini Live API without traditional speech-to-text pipelines. Cox Automotive integrated the technology into an AI-powered shopping assistant for Autotrader, using live screen-highlighting and tool-calling to guide car shoppers through searches, comparisons, and financing via natural conversation.

Customer testimonials underscore the practical impact of Gemini 3.8 Live. Equal AI reported handling over a million live calls daily across nine Indian languages, with improvements in interruption handling, multilingual conversations, and tool-call reliability. Salesforce AI Research highlighted collaboration with Google to combine real-time multimodal capabilities with agentic AI for richer customer experiences. Specs noted significant updates to Voice Activity Detection and latency reductions during testing and benchmarking.