GUIDE

React Native text to speech: from system voices to your own voice server

A React Native app can speak in three ways: with the voices of the phone, with audio files voiced in advance, or with a speech server that voices new text on demand. The phone’s voices are free and offline but sound different on every device; the other two give your app one recognizable voice.

SuperClonica gives you natural voices in 31 languages, free to try online, a free engine to voice the app’s lines on your computer and a Docker server for text that changes. This guide shows each approach in React Native and Expo.

Browse voices Install the SuperClonica engine

Three ways to make an app speak

SuperClonica speaks 31 languages; Chinese is not supported yet, so a Chinese interface still needs another voice.

  • expo-speech or react-native-tts: the voices installed on the phone, free and offline, different on every device, and they cannot save audio.
  • Pre-voiced assets: every fixed line voiced once with SuperClonica and shipped in the app.
  • A speech server: the app sends text to your SuperClonica server and plays the answer.
  • Most apps mix them: assets for fixed lines, a server for user text, the system voice as a fallback.

Pick a voice free online

On the page of any voice, type a line of your app, 30 to 100 characters, choose the language and press Voice it. In a few seconds you hear whether it fits your app, with no account, also on your phone.

Pre-voiced lines with expo-audio

Name the files by the key you already use for translations, so every language has the same file names in its own folder.

  • Install the free engine on your computer: python -m pip install superclonica-cpu.
  • Voice each line: superclonica --text "Welcome back!" --voice F1 --lang en --output assets/voice/welcome.mp3.
  • Play it with expo-audio: const player = useAudioPlayer(require('./assets/voice/welcome.mp3')); player.play();
  • Assets work offline and add no delay; MP3 keeps a one-sentence line to a few dozen kilobytes.

A speech server for user text

For text that changes, such as chat messages, user notes or a tutor that answers, run the server on a machine you control: docker run -d -p 8000:8000 -v superclonica-models:/models superclonica/superclonica:cpu. Put it behind HTTPS and set SUPERCLONICA_API_KEY so only your app can call it. The app posts JSON with text, voice and lang to /synthesize and gets a WAV back; save it to a file with your file-system library and give the file to the audio player. Cache answers for texts that repeat: the same text, voice and seed always give the same file.

When the system voice is enough

The system voice is enough when the voice only has to be understood: reading a notification aloud, an accessibility helper or a prototype that must work without any network. Choose your own voice when it is part of the product, such as a tutor, a game character or a guided meditation. For an app you publish, use the ten stock voices, M1 to M5 and F1 to F5, free under the OpenRAIL-M license, or your own voice trained in the SuperClonica studio; fan-made character voices are not for store apps.