Features
Speech Generation
Generate speech through OffRail's OpenAI-compatible audio route.
Speech Generation
Create speech with POST /v1/audio/speech. Choose the current speech model and
voice from the public catalogue.
Request
curl "https://api.offrail.ai/v1/audio/speech" \
-H "Authorization: Bearer $OFFRAIL_API_KEY" \
-H "Content-Type: application/json" \
-d "{\
\"model\": \"$OFFRAIL_MODEL\",\
\"input\": \"Say what should be spoken.\",\
\"voice\": \"$OFFRAIL_VOICE\"\
}" \
--output speech.wav| Field | Required | Description |
|---|---|---|
model | Yes | Public speech model ID |
input | Yes | Text to synthesize |
voice | Yes | Voice supported by the selected model |
response_format | No | Output format when the selected model supports it |
The response body contains audio. Use the returned Content-Type instead of
assuming a fixed format.
Speech pricing is model-specific. The public model page states whether usage is billed by character, token, second, or request.
See the speech API reference for the complete request and response schema.