Text to speech

Synthesize speech from text with preset, designed, or cloned voices.

Text to speech takes a string and returns spoken audio through an API that follows the OpenAI speech shape. Pick a voice, send the text, and the response body is the audio itself. If you already call OpenAI for speech, change the base URL and the model.

Get an API key

Create a Sprag account and open API keys in the app to generate one.

Copy the key immediately. The full value is shown once and is not recoverable after you leave the page.

Export it as an environment variable, replacing the placeholder with your key:

export SPRAG_API_KEY=<your_sprag_api_key>

Make a request

Point any OpenAI-compatible client at https://api.sprag.ai/v1 and set model to chorus. The response body is the audio itself, not JSON.

curl https://api.sprag.ai/v1/audio/speech \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $SPRAG_API_KEY" \
  -d '{
    "model": "chorus",
    "input": "The quarterly numbers are in and they look good.",
    "voice": "marin",
    "response_format": "mp3"
  }' \
  --output out.mp3

Pick a voice

marin above is a preset voice. List the voices rather than hardcoding an id. The set can change.

curl "https://api.sprag.ai/v1/audio/voices?model=chorus" \
  -H "Authorization: Bearer $SPRAG_API_KEY"

A voice comes from one of three sources, and the fields you send choose which:

Send	Voice source
`voice` with a preset id	A built-in speaker
`voice` with the id of a designed voice	A voice created from a written description
`ref_audio` and `ref_text`, or `voice` with the id of a cloned voice	A voice cloned from reference audio

Design happens before the speech request. POST /v1/audio/voices/design returns previews, and POST /v1/audio/voices creates a voice from one. The voice's id goes in voice. Voices covers all three sources.

Where to go next

API. REST for text you already have, realtime for spoken conversation.

Voices. Presets, designed voices, and cloning.

Models. The speech models and the fields each one accepts.

Concepts. Language coverage and audio formats.