stts 1.2.4

Speech-to-Text and Text-to-Speech plugin. Offline first.

stts

Speech-to-Text (STT) and Text-to-Speech (TTS) Flutter plugin.

No dependency. All implementations use what the platform provides.

Platform Speech-to-Text parity matrix

Feature	Android	iOS	macOS	web	Windows
permission	✔️	✔️	✔️	✔️
language selection	✔️	✔️	✔️	✔️	✔️
punctuation	✔️	✔️	✔️
grammars	✔️*	✔️	✔️	✔️*

• *: seems to do nothing.
• Specific platform features are not listed here.

Platform Text-to-Speech parity matrix

Feature	Android	iOS	macOS	web	Windows
pause/resume	✔️	✔️	✔️	✔️	✔️
language selection	✔️	✔️	✔️	✔️	✔️
voice selection	✔️	✔️	✔️	✔️	✔️
pitch	✔️	✔️	✔️	✔️	✔️
rate	✔️	✔️	✔️	✔️	✔️
volume	✔️	✔️	✔️	✔️	✔️

Usage of Speech-to-Text

import 'package:stts/stts.dart';

final stt = Stt();

// Get state changes
final sub = stt.onStateChanged.listen(
  (speechState) { /* SttState.start/stop */ },
  onError: (err) {/* Retrieve listener errors from here */ },
);

await tts.hasPermission(); // Request for audio recording and STT service usage.

// Get intermediate and final results.
stt.onResultChanged.listen((result) { /* The current result String */ });

// Start speech recognition.
stt.start(); // ...optionnaly, abort with stt.stop();

// As always, don't forget to release resources.
sub.cancel();
stt.dispose();

Usage of Text-to-Speech

import 'package:stts/stts.dart';

final tts = Tts();

// Get state changes
final sub = tts.onStateChanged.listen(
  (ttsState) { /* TtsState.start/stop/pause */ },
  onError: (err) { /* Retrieve listener errors from here */ },
);

// Add utterance. Texts are queued.
tts.start('Hello');
tts.start('world!'); // ...optionnaly, abort with tts.stop();

// As always, don't forget to release resources.
sub.cancel();
tts.dispose();

Platforms setup & infos

You can either use one or both engines in your app. So permissions are only required for dedicated engine.

• Android
• iOS
• macOS
• Web
• Windows

Misc. infos / warnings

• Speech recognition is time limited. You can't use it to fill very long texts in a single session.
• Speech recognition will auto stop, when detecting silence.
• This plugin is offline first, this can't be guaranteed.