KittenTTS React Native

On-device text-to-speech for React Native.
Generate speech on iOS and Android without sending text to a cloud TTS API.

Developer preview. APIs may change between releases.

Expo Go will not work. KittenTTS uses native modules (onnxruntime-react-native and react-native-fs), so Expo apps need a development build or a prebuilt native project.

See It In Action

Device: iOS · Expo example     Device: Android · Word timings

---

What Is KittenTTS React Native?

KittenTTS React Native lets you add local speech synthesis to a mobile app:

• Text-to-speech - neural voice synthesis from plain text.
• On-device inference - powered by KittenTTS and ONNX Runtime.
• Private by default - no cloud TTS request after assets are available.
• Offline-ready - download once into the device cache, or bundle assets into the native app.
• App-friendly output - play audio directly, save WAV data, or use generated word timings for read-aloud UI.

No cloud. No API key. No text leaving the device for speech generation.

---

SDK

Runtime	Status	Docs
React Native iOS	Developer preview	Getting started
React Native Android	Developer preview	Getting started
Expo development build	Supported	Expo setup
Expo Go	Not supported	Why not?

Install:

npm install @kittentts/react-native

---

Quick Start

Install the SDK:

npm install @kittentts/react-native

Generate a WAV in memory:

import { KittenTTS } from '@kittentts/react-native';

const tts = await KittenTTS.create(undefined, (progress) => {
  console.log(`setup ${Math.round(progress * 100)}%`);
});

const result = await tts.generate('Hello from KittenTTS on React Native.');

console.log(result.duration);
console.log(result.wavBase64());

await tts.dispose();

Play audio through Expo Audio:

import * as ExpoAudio from 'expo-audio';
import { KittenTTS, createExpoAudioPlayer } from '@kittentts/react-native';

const tts = await KittenTTS.create({
  player: createExpoAudioPlayer(ExpoAudio),
});

await tts.speak('This voice is generated on the device.');

Full getting started guide →

---

Expo Setup

Expo Go cannot load the native ONNX Runtime and filesystem modules used by this SDK. Use a development build:

npm install @kittentts/react-native
npx expo install expo-audio expo-dev-client
npx expo prebuild
npx expo run:ios

For Android:

npx expo run:android

After the development build is installed:

npx expo start --dev-client

If the app opens in Expo Go, stop it and run npx expo run:ios or npx expo run:android again.

---

Sample Apps

• examples/BareRNExample - bare React Native setup.
• examples/ExpoExample - Expo development build setup.
• examples/ExpoWordTimingsExample - word highlighting with generated timings.
• examples/OfflineBundledAssetsExample - shipping models and phonemizer files inside the app.

---

Features

• On-device TTS inference on iOS and Android.
• Model download and cache with progress callbacks.
• Bundled offline assets for apps that cannot depend on a first-run download.
• Expo development builds; Expo Go is not supported.
• Playback helpers for Expo Audio, React Native Sound, and custom audio layers.
• WAV output as raw PCM samples, bytes, or base64.
• Word timings for read-aloud highlighting.
• Streaming generation for longer text.

---

Supported Models

Start with NanoInt8 for the smallest download. Use larger models when quality matters more than size.

Model	Enum	Parameters	Approx download	Use case
Nano int8	KittenModel.NanoInt8	15M	28 MB	Smallest app/download size
Nano fp32	KittenModel.Nano	15M	59 MB	Nano quality without quantization
Micro	KittenModel.Micro	40M	44 MB	Better quality, still compact
Mini	KittenModel.Mini	80M	83 MB	Highest quality option

Models and voices →

Voices

Bella, Jasper, Luna, Bruno, Rosie, Hugo, Kiki, Leo

await tts.speak('Luna speaking.', KittenVoice.Luna);
await tts.speak('Slower Bruno speaking.', KittenVoice.Bruno, 0.85);

---

Docs

• Docs overview
• Getting started
• Playback
• Bundled offline assets
• Word timings
• Models and voices
• API reference
• Troubleshooting

Tutorials:

• Simple TTS app on macOS
• Simple TTS app on Windows
• Reader app on macOS
• Reader app on Windows

---

System Requirements

• React Native >= 0.72
• iOS 15.1+
• Android API 24+
• Node.js 20+ recommended for examples

Runtime dependencies installed by the SDK:

• onnxruntime-react-native
• react-native-fs
• pako

Audio playback is optional. Use expo-audio, react-native-sound, or a custom player.

---

Roadmap

• Improve streaming and playback queue primitives.
• Add more reader-app helpers around pause, resume, and seek state.
• Continue tracking ONNX Runtime compatibility across React Native and Expo releases.
• Support future KittenTTS model releases as they become available.

Need something specific? Open an issue.

---

Community And Support

• Website: kittenml.com
• Repository: KittenML/kittentts-react-native
• Discord: Join the community
• Demo: Hugging Face Spaces
• Issues: GitHub Issues
• Commercial support: contact form

License

Apache 2.0. See LICENSE.