Interface AudioStreamScript

A script that makes the agent lip-sync an audio file you host, with no text-to-speech involved.

The other payload speak() accepts. Use it when the audio already exists (a recording, or speech you synthesized yourself) instead of having a provider generate it from text. Talks (V2) and Clips (V3) agents only: an Expressive (V4) agent drops an audio script without an error.

Example: Audio file

const speak = await agentManager.speak({
    type: 'audio',
    audio_url: 'https://www.yourwebsite.com/audio.mp3',
});

interface AudioStreamScript {
audio_url: string;
type: "audio";
}

Index

Properties

audio_url type

Properties

audio_url

audio_url: string

URL of the audio file the agent lip-syncs to.

The URL has to be publicly reachable, since the file is fetched server-side rather than uploaded from the browser. The file may be at most 15 MB.

Example

"https://www.yourwebsite.com/audio.mp3"

type

type: "audio"

Which kind of script this is. Always audio for this variant; a TextStreamScript carries text instead.

Interface AudioStreamScript

Example: Audio file

Index

Properties

Properties

audio_url

Example

type

Settings

On This Page