Generate a recording from text

Turns text into speech and saves it as a WAV recording on your account. Each call is billable. A 201 confirms the recording is saved, not that it is already playable everywhere: delivering the audio to the phone system continues after the response, and a failure at that step is not reported back to you.

Recent Requests
Log in to see full request history
TimeStatusUser Agent
Retrieving recent requests…
LoadingLoading…
Body Params
string
required

Name of the recording, also used to build its audio filename. Required; a missing name is a 400.

string

Description of the recording. Defaults to TTS: followed by the first 100 characters of tts.input.

tts
object
required

What to speak and how. tts.input is required.

string
required

The text to speak. Required.

string

Speech model: gemini-3.8-flash-tts (default) or gemini-3.8-flash-lite-tts. Any other value is a 400 with code tts_model_invalid, and the response's valid_models lists the accepted values.

string

Voice ID to speak in. Defaults to fr-ca-csagent-5. An unknown ID is a 400 with code tts_voice_invalid, and the response's valid_voices lists every voice you can use.

string

Plain-language direction for how the text is spoken (tone, pace, clarity, how to pronounce particular names), up to 500 characters. Defaults to Professional, slow, and enunciate clearly. Longer text is a 400 with code tts_style_invalid.

Responses
201

The recording was generated and saved.

401

Missing or invalid API key.

Language
Credentials
Bearer
LoadingLoading…
Response
Click Try It! to start a request and see the response here! Or choose an example:
application/json