Chunks are scheduled to the Web Audio API as they arrive — playback starts on the first chunk, not after the last.
Not connected
Click a built-in voice or type a custom HuggingFace URL (hf://…) or local path.
First use will download it automatically.