Loading video player...
Build a voice agent with Gemini 3 Flash Live and LiveKit. Jesse Hall from LiveKit walks through everything you need to get a working voice agent running locally, then take it to production. Speech-to-speech audio, no text in the middle, with better instruction following, smarter tool calling, and reduced speaker drift across long sessions. What's covered: Project setup with LiveKit Agents and the Google plugin, system prompt best practices for voice, native audio capabilities, server-side and client-side tool calling, Google Search integration, and live testing with multilingual switching and function chaining. Grab your Gemini API key at Google AI Studio and your LiveKit credentials at cloud.livekit.io to get started. Resources: ✅Livekit Gemini Playground → https://goo.gle/3NpT0OU ✅Gemini Hacker Starter Repo → https://goo.gle/4dMdxrr ✅GitHub examples → https://goo.gle/4tfKKjM What are you building with Gemini Live? Drop it in the comments. Subscribe to Google for Developers → https://goo.gle/developers Speaker: Jesse Hall from LiveKit Products Mentioned: Google AI, Gemini