Conversational AI
Coming soon: interactive voice and avatar-style real-time conversations. The live conversation backend is not deployed yet, so this page describes the planned experience rather than one you can open today.
Conversational AI has not shipped yet, so there is nothing to open. This page documents the planned experience — the related guides below cover what you can use today.
Overview
Conversational AI will be the interactive surface for real-time assistant experiences — live speaking and response flows that go beyond one-shot generation. It is not available yet. For conversational workloads you can run today, use Agents Platform for voice agents and call flows, or Streaming for low-latency text-to-speech playback.
Quickstart
Use Agents Platform in the meantime
Build and monitor voice agents, live calls, and workflow-driven conversations at /app/agents-platform — that is the shipped path for real-time conversational work.
Stream low-latency speech today
If you only need incremental audio playback rather than a full duplex conversation, Streaming already delivers chunked text-to-speech.
Watch this page for the release
This guide is updated with real setup steps when the live conversation backend is deployed.