---
title: "Efficient On-Device LLMs in React Native Done Right"
description: "Artur Morys-Magiera will cover efficient on-device LLMs in React Native, including runtimes, model selection, quantization, and platform constraints."
url: https://www.callstack.com/events/efficient-on-device-llms-in-react-native-done-right
updated: 2026-10-24T08:39:19.590Z
---

# Efficient On-Device LLMs in React Native Done Right

Artur Morys-Magiera will cover efficient on-device LLMs in React Native, including runtimes, model selection, quantization, and platform constraints.

_Talk · 11/17/2026 · New York, United States · Online · Category: AI_

Organizer: React Summit

Speakers: Artur Morys-Magiera (Software Engineer @ Callstack)

[Event page](https://gitnation.com/events/react-summit-us-2026)

Running LLMs on a mobile device moves model size, memory use, hardware acceleration, runtime choice, and platform variability into the app architecture.

In this talk, Artur will explain how these constraints affect latency, privacy, reliability, and user experience across iOS and Android. The session compares practical runtime and model options available to React Native applications, including system-provided models and application-managed inference.

It also covers optimization choices such as quantization, model selection by device capability, and a shared API for local and remote inference.