Create Custom Voice
Create a custom voice in your workspace from a voice embedding.
Creates the voice with Cartesia and adds it to your workspace. Get the embedding from Clone Voice first.
To use the voice, set it on a call agent as ttsModel: { "provider": "cartesia", "voiceId": "<id>" }, using the id from the response.
Headers
x-workspace-idstringrequiredWorkspace identifier for the API.
Request Body
namestringrequiredName for the custom voice.
embeddingarrayrequiredThe embedding returned by Clone Voice, sent unchanged.
descriptionstringDescription of the custom voice.
languagestringdefault: enLanguage code for the voice (e.g., "en").
Response
200 with the voice object Cartesia returns, passed through unchanged.
idstringThe voice ID. Use it as ttsModel.voiceId (with provider cartesia) and to delete the voice.
namestringName of the custom voice.
descriptionstringDescription of the custom voice.
languagestringLanguage code of the voice.
created_atstringTimestamp when the voice was created.
Errors
| Status | Body | When |
|---|---|---|
400 | {"error": "Workspace ID is required"} | No x-workspace-id |
400 | {"error": "Name and embedding are required"} | name or embedding is missing |
500 | {"error": "..."} | Cartesia refused the request (for example a malformed embedding), or anything else went wrong |
{
"id": "aaaaaaaa-bbbb-4ccc-8ddd-eeeeeeeeeeee",
"name": "Custom Sales Voice",
"description": "Professional sales voice",
"language": "en",
"created_at": "2026-10-29T09:00:00.000Z"
}Last updated on