Skip to main content

On-device AI for React Native

Run LLMs, computer vision, and speech models directly on mobile devices with a declarative React Native API, powered by ExecuTorch.

Get StartedAPI Reference
// Run Llama 3.2 on device
const llm = useLLM(
model: LLAMA3_2_1B
await llm.generate(chat);

Quick start

Get running in minutes with our step-by-step guide

1

Install the package

Add React Native ExecuTorch to your project with your preferred package manager.
npm install react-native-executorch
2

Add resource fetcher

Install the appropriate resource fetcher for your setup (Expo or bare React Native).
yarn add @react-native-executorch/expo-resource-fetcher
yarn add expo-file-system expo-asset
3

Initialize and use

Import the library, initialize with your resource fetcher, and start using AI models.
import {
  useLLM,
  LLAMA3_2_1B,
  initExecutorch,
} from 'react-native-executorch';
import { ExpoResourceFetcher } from '@react-native-executorch/expo-resource-fetcher';

initExecutorch({
  resourceFetcher: ExpoResourceFetcher,
});

function MyComponent() {
  const llm = useLLM({ model: LLAMA3_2_1B });
  
  const handleGenerate = async () => {
    const chat = [
      { role: 'system', content: 'You are a helpful assistant' },
      { role: 'user', content: 'What is the meaning of life?' }
    ];
    await llm.generate(chat);
    console.log('Response:', llm.response);
  };
  
  return (
    <Button onPress={handleGenerate} disabled={!llm.isReady}>
      Generate
    </Button>
  );
}

Explore capabilities

Choose the AI capability that fits your use case

Large Language Models

Run Llama and Qwen models on-device with tool calling, chat history, and context strategies

Computer Vision

Image classification, object detection, segmentation, OCR, style transfer, and embeddings

Speech Processing

Speech-to-text, text-to-speech, and voice activity detection for audio applications

Text Embeddings

Generate semantic embeddings for search, similarity, and recommendations

Key features

Everything you need to build on-device AI applications

Declarative React API

Use familiar React hooks to integrate AI models without native code expertise

100% On-device

All inference runs locally on iOS and Android devices for privacy and offline use

Ready-made models

Pre-optimized models from HuggingFace ready to use with simple constants

Optimized loading

Automatic model downloading, caching, and progress tracking built-in

Resources & community

Get help and connect with other developers

GitHub Repository

View source code, report issues, and contribute to the project

Discord Community

Join our Discord server to ask questions and share your projects

HuggingFace Models

Browse our collection of optimized models ready for on-device use

Private Mind App

See React Native ExecuTorch in action in a production privacy-first AI app

Ready to build with on-device AI?

Start building privacy-first mobile AI applications with React Native ExecuTorch today.

Get Started Now