mstar.api_server.openai.serving_speech

/v1/audio/speech handler (text-to-speech).

Non-streaming returns the full audio as a container blob (WAV by default). Streaming returns a single open-ended WAV response (header + PCM16 frames) as the audio is produced.

Functions

create_speech(api, model_name, adapter, req)

async mstar.api_server.openai.serving_speech.create_speech(api, model_name, adapter, req)