TTS Commands

Commands for generating narration audio with word-level timing.

synthesize

Synthesizes narration text to audio with word-level timing events.

text

string

required

Plain text to synthesize

bundle_path

string | null

Path to TTS bundle directory for caching. If provided and a cache hit occurs, returns cached audio immediately.

on_event

Channel<TTSEvent>

required

Event channel for streaming word boundaries and audio data

result

void

required

Emits events via channel:

WordBoundary: Fired for each word
AudioReady: Fired once when synthesis completes

TTSEvent

Discriminated union:

Show WordBoundary

event

'wordBoundary'

required

data.word

string

required

Word text

data.wordIndex

number

required

Zero-based word index

data.charOffset

number

required

Character offset in original text

data.startMs

number

required

Word start time (milliseconds)

data.endMs

number

required

Word end time (milliseconds)

Show AudioReady

event

'audioReady'

required

data.audioBase64

string

required

Base64-encoded WAV audio

data.durationMs

number

required

Total audio duration (milliseconds)

import { Channel } from '@tauri-apps/api/core';

const channel = new Channel<TTSEvent>();
channel.onmessage = (event) => {
  if (event.event === 'wordBoundary') {
    console.log(`Word: ${event.data.word} at ${event.data.startMs}ms`);
  } else if (event.event === 'audioReady') {
    const audio = new Audio(`data:audio/wav;base64,${event.data.audioBase64}`);
    audio.play();
  }
};

await invoke('synthesize', {
  text: 'Hello, world!',
  bundlePath: '/path/to/bundle',
  onEvent: channel
});

export_audio

Pre-generates and caches audio for multiple text blocks.

texts

string[]

required

Array of text blocks to synthesize

bundle_dir

string

required

Path to bundle directory for writing cache

exported

number

required

Number of newly generated audio files (cache misses)

const exported = await invoke<number>('export_audio', {
  texts: ['Intro text', 'Section 1', 'Section 2'],
  bundleDir: '/course/tts-bundle'
});

console.log(`Generated ${exported} new audio files`);

ensure_tts_ready

Ensures TTS models are downloaded and ready. Performs a warmup synthesis if models need initialization.

status

string

required

One of:

'ready': Models already present
'downloaded': Models downloaded and initialized

const status = await invoke<string>('ensure_tts_ready');
if (status === 'downloaded') {
  console.log('TTS models initialized');
}

This command is called automatically by the app on startup. You typically don’t need to invoke it manually.

Type Definitions

Tauri Commands

synthesize

TTSEvent

export_audio

ensure_tts_ready

Build docs developers (and LLMs) love

Type Definitions

Tauri Commands

​synthesize

​TTSEvent

​export_audio

​ensure_tts_ready

Build docs developers (and LLMs) love

synthesize

TTSEvent

export_audio

ensure_tts_ready