Uploaded February 2026 | Updated September 2026, 10 hours ago

---

Running LLMs on-device changes how mobile AI applications are built.

In this talk from React Advanced London, Szymon Rybczak demonstrates how to run large language models directly inside React Native apps using react-native-ai.

You’ll learn:

Why on-device LLMs matter for privacy, latency, and offline use
How react-native-ai abstracts local AI execution in React Native
How the provider architecture works
Integrations with MLC LLM Engine and Apple’s foundation models
What’s new: better debugging, tool calling, and agent pipelines

The session focuses on real implementation details and trade-offs, giving you a clear picture of what’s possible today when AI runs entirely on the device.

Check out more AI insights: clstk.com/3ZrOnWW