Documentation Index
Fetch the complete documentation index at: /llms.txt
Use this file to discover all available pages before exploring further.
Generate speech through VoidAI with supported text-to-speech models, voices, speed controls, and binary audio responses.
POST https://api.voidai.app/v1/audio/speech
Requires authentication and a model that advertises this endpoint.
| Field | Type | Description |
|---|---|---|
model | string, required | A supported speech model, such as tts-1. |
input | string, required | Text to speak. Provider limits may apply. |
voice | string, required | alloy, echo, fable, onyx, nova, or shimmer. |
response_format | string | mp3, opus, aac, flac, wav, or pcm; provider support varies. |
speed | number | Between 0.25 and 4.0. |
curl https://api.voidai.app/v1/audio/speech \
-H "Authorization: Bearer $VOIDAI_API_KEY" \
-H 'Content-Type: application/json' \
-d '{"model":"tts-1","input":"Hello from VoidAI.","voice":"alloy","response_format":"mp3"}' \
--output speech.mp3
audio/mpeg with the filename speech.mp3, including when another format is requested; do not use that header alone to infer the encoding. Check models and pricing before making requests.