Skip to main content
Deepgram's Docs

Search documentation

Type to search this documentation.

On this pageOverview

Agent Audio Done

Voice Agent

The server sends AgentAudioDone immediately after the last audio chunk for an agent utterance. Use this event to mark the end of a server-side stream so you can synchronize follow-up actions with the agent finishing.

AgentAudioDone does not mean the user has heard the agent finish. Audio you already received may still sit in your local playback buffer. To detect end-of-playback in the user's ears, watch your audio output queue, not this event.

JSON

JSON
{
  "type": "AgentAudioDone"
}
Suggest an edit

Propose a replacement for this page. The site team reviews it before applying any changes.

Export
Documentation menu