The Generative UI SDK for React Native
Built for bare React Native CLI projects. Stop piping LLM text into static views, give your agents the power to render native mobile components with type-safe props, straight on the Hermes runtime, no prompt engineering and no Expo required.
Made for the CLI workflow
WireAI is a plain npm dependency, so it slots into a project scaffolded with the React Native Community CLI without changing your native build setup. Keep your existing Xcode and Gradle config, install, wrap your root component, and your agent can render UI.
Tuned for Hermes
Generated UI is rendered with standard React Native primitives on the Hermes engine, not a WebView. Register a component, return validated props from the model, and you get the same native performance as the rest of your bare RN app.
SDK Features
Open Source Today
WireAI is MIT licensed and ready to drop into your next bare React Native CLI project.
SDK FAQ
Does WireAI work in a bare React Native CLI project?
Yes. WireAI installs as a regular npm dependency, so it drops into a project created with the React Native Community CLI the same way it does an Expo project. Wrap your app in the provider and register your components. No Expo modules are required.
How does WireAI differ from a React Native boilerplate?
WireAI is an SDK, not a boilerplate. A boilerplate gives you a starting file structure; WireAI gives you a runtime for generative UI. You add it to any existing React Native project to let an AI agent render native components within minutes.
Can I use WireAI with my own custom components?
Yes. That is the core feature. You register your existing React Native components with a name, description, and Zod schema. WireAI then lets your AI agent 'wire' those components into the conversation with validated props.
Does it support local LLMs like Ollama?
Absolutely. WireAI is built to work offline and locally. You can point it at an Ollama or LM Studio endpoint and run generative UI entirely on your own machine or device without an API key.