Skip to main content
List every voice your account can use, or retrieve one by voice_id. Pass the voice_id when you generate speech.
A Voice object describes one voice. Pass its voice_id when you use batch or open a realtime reply. See Voices and languages for the full catalog.

The Voice Object

Attributes

string
Unique identifier for the voice.
string
Display name.
string
Values: male, female.
array
Language codes the voice speaks, for example ["en-US"] or ["hi"]. Each voice speaks one language today.
string
The voice’s accent: American, British, Australian or Indian.
string
A short description of the voice.
boolean
Whether the voice supports custom pronunciations: inline pronunciations or a saved lexicon passed as lexicon_id. Which rule kinds it accepts (ipa, arpa, respell) depends on the voice; a kind it doesn’t take returns invalid_request.
string
URL of a short audio sample of the voice.

Example Voice Object


List All Voices

List the voices your account can use.

Arguments

No arguments. Query parameters aren’t accepted: one returns 422 invalid_request naming it.

Response

string
Unique identifier for the request.
object
limit, offset, total_count, next and previous. All voices come back in one page.
array

Retrieve a Voice

Get one voice by its voice_id.

Arguments

No arguments. A voice_id that isn’t in your list returns 404 voice_not_found.

Response

The Voice object, with the request’s api_id.