Text to Speech
Synthesize natural speech from text, with streaming output.
Kotoba’s Text-to-Speech (TTS) synthesizes audio from text and streams it back as soon as the model produces it, so you can pipe straight to a speaker, a WebRTC track, or an LLM-driven agent without waiting for the full utterance.
Supported language: Japanese (ja).
Three call shapes
- One-shot — pass a full string in, get a complete waveform back
(
pcm_f32@ 24 kHz by default; choosepcm16,mulaw, oropuswithaudio_format=). - Streaming — send the text in one frame and read audio chunks as they emerge from the server (output streaming).
- Async — the same surface, with
AsyncKotobaClientfor production-grade concurrency.
Available Japanese speakers: male-01 (default) and female-01.
voice_settings adjusts the speed, the pitch and the volume of a voice, for
a whole session or for one response.
Where to go next
- Python SDK for TTS —
client.tts.synthesize(...)for one-shot,client.tts.stream(...)for streaming. Both acceptspeaker_id,audio_format,sample_rate, andvoice_settings. - API reference — the AsyncAPI spec for the TTS WebSocket channel.