Flutter SDK

The kittentts package generates local speech across Flutter targets. Native platforms use ONNX Runtime through flutter_onnxruntime; web uses ONNX Runtime Web and a bundled JavaScript CE phonemizer.

Package 0.1.1Flutter 3.22 and Dart 3.4 or later

Broad platform setAndroid, iOS, macOS, Linux, Windows, and web

App-owned playbackGenerate PCM/WAV and plug in the audio layer you already use

import 'package:kittentts/kittentts.dart';

final tts = await KittenTTS.create();
final result = await tts.generate('Hello from KittenTTS on Flutter.');
print(result.duration);
print(result.wavBase64());
await tts.dispose();

Capabilities

Four local models and eight English voices.
In-memory generation, long-text streams, and word timings.
Custom player interface and FIFO playback queue.
Model cache inspection, predownload, validation, redownload, and clear helpers.
Local paths or Flutter web asset and HTTP URLs.
Anonymous generation analytics with a per-session opt-out.

Repository: KittenML/KittenTTS-flutter

Capabilities​

Capabilities