Generate Speech
Convert text to speech with the OpenAI-compatible audio API.
Generate streamed audio from text using a supported speech model and voice.
POST
/openai/v1/audio/speechRequest Body
| Field | Type | Required | Description |
|---|---|---|---|
model | string | Yes | Text-to-speech model identifier. |
input | string | Yes | Text to synthesize. |
voice | string | Yes | Voice supported by the selected provider. |
response_format | string | No | Requested audio format, such as mp3. |
Response
The endpoint streams binary audio. Use the response content type to choose the correct file extension or playback decoder.
Transcription
Use POST /openai/v1/audio/transcriptions with multipart form data. Both file and model are required. Do not rely on a default transcription model; choose a current model from the published catalog.