What if your React Native app could run a large language model right on the device, with no cloud required? That’s no longer a hypothetical.

In this live stream, Callstack’s Mike Grabowski, Szymon Rybczak, and Kewin Wereszczński walk you through how they made on-device LLM execution possible using the MLC LLM Engine and the new react-native-ai package.

The project started as a window.ai-inspired proof of concept showcased at The Geek Conf. Now, it’s a growing open-source effort under the Callstack Incubator, aimed at making edge AI development easier and more accessible for mobile devs. You’ll see how the integration with Vercel’s AI SDK simplifies model switching, how MLC powers real-time text generation locally, and what we’re doing next to expand model support and improve developer experience.

Expect a hands-on walkthrough, straight talk about implementation trade-offs, and a peek at what’s coming, from better documentation to deeper platform integration. Whether you’re building for privacy, performance, or offline resilience, this session will show you where AI in React Native is headed—and how you can be part of it.