Practical overview
When it fits
Consider it when building a React Native or Expo app that needs on-device transformer inference and you can accommodate its native setup and the repository's inactive maintenance status.
Who it is for
- React Native developers adding local language-model inference to mobile apps.
- Expo developers who can configure native dependencies and use a development client.
Documented uses
- Generate text in a React Native or Expo app using an on-device transformer model.
- Build app interactions that answer questions or process language without sending prompts to a hosted model service.
- Download and cache model files for use by an app.
Requirements & limits
- Inference uses ONNX Runtime through the onnxruntime-react-native dependency.
- Model files must be fetched or otherwise made available to the app; the README shows a fetch callback and an Expo cache example.
- Expo development requires a development client rather than Expo Go because of native code.
- The repository notice says active maintenance ended in July 2025.
Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.
Set up the library in an Expo app
Configure the documented native dependencies and Babel plugin for an Expo integration.
Before you start
- An Expo application
- onnxruntime-react-native
- A development client using EAS Development Build or Expo Prebuild
Steps
Install onnxruntime-react-native and react-native-transformers.
Add onnxruntime-react-native to the Expo plugins configuration in app.json or app.config.js.
Add babel-plugin-transform-import-meta to the Babel plugins configuration.
Use a development client rather than Expo Go because the library uses native code.
Expected outcome
The documented Expo-side dependency and configuration setup is in place for integrating the library.
Guide limitations
- This setup recipe does not cover loading a model or generating text.
- The repository says the project is no longer actively maintained.
Load an ONNX Llama model and generate text
Follow the README example to initialize a Hugging Face ONNX model and pass generated text to an app interface.
Before you start
- A configured React Native or Expo app with the library and ONNX Runtime dependency
- A model repository and ONNX model file
- A fetch function that returns the downloaded model file location
Steps
Import Pipeline from react-native-transformers.
Initialize Pipeline.TextGeneration with the model repository and ONNX file path.
Provide the required fetch function for model files.
Call Pipeline.TextGeneration.generate with a prompt and use its callback to update the displayed output.
Expected outcome
The example connects a text-generation prompt to generated text displayed in the app.
Guide limitations
- The README example downloads model files; it does not establish that they are already available offline.
- The repository says the project is no longer actively maintained.
Sources & review
Source record details
- Source availability
- Active
- Sources checked
- 15 October 2026
- Original website title
- daviddaytw/react-native-transformers · Official homepage title · retained in its source language