Create Audio Generation Request

Generate audio from input text

POST /audio/speech

Create audio generation request

cURL

curl --request POST \
  --url https://api.together.xyz/v1/audio/speech \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '
{
  "model": "cartesia/sonic",
  "input": "<string>",
  "voice": "laidback woman",
  "response_format": "wav",
  "language": "en",
  "response_encoding": "pcm_f32le",
  "sample_rate": 44100,
  "stream": false
}
'

200

"<string>"

Authorizations

Authorization · string · header · default: default · required

Bearer authentication header of the form Bearer <token>, where <token> is your auth token.

Body

application/json

model · enum<string> string · required

The name of the model to query.

See all of Together AI's chat models

Available options: cartesia/sonic

Example:

"cartesia/sonic"

input · string · required

Input text to generate the audio for

voice · enum<string> string · required

The voice to use for generating the audio. View all supported voices here.

Available options: laidback woman, polite man, storyteller lady, friendly sidekick

response_format · enum<string> · default: wav

The format of audio output

Available options: mp3, wav, raw

language · enum<string> · default: en

Language of input text

Available options: en, de, fr, es, hi, it, ja, ko, nl, pl, pt, ru, sv, tr, zh

response_encoding · enum<string> · default: pcm_f32le

Audio encoding of response

Available options: pcm_f32le, pcm_s16le, pcm_mulaw, pcm_alaw

sample_rate · number · default: 44100

Sampling rate to use for the output audio

stream · boolean · default: false

If true, output is streamed for several characters at a time instead of waiting for the full response. The stream terminates with data: [DONE]. If false, return the encoded audio as octet stream

Response

200

application/octet-stream

OK

The response is of type file.