Uploaded June 2026 | Updated September 2026, 1 week ago
Models designed to run on edge devices are more popular than ever, and running them on phones with React Native seems straightforward, but it introduces a unique set of challenges you won't see coming. This talk draws on a year and a half of building react-native-executorch and shipping real applications with on-device inference. We'll walk through the problems that surface only once you actually try: app sizes exploding, models crashing on mid-range devices, batteries draining during continuous inference, and the UX challenges of long model loads. We'll cover both LLM and computer vision use cases, from on-device chat to real-time object detection, and the very different constraints each brings to mobile. You'll leave knowing which mobile AI use cases are production-ready today and which ones still need more time.
Don't miss any updates about App.js Conf and follow us on X:
💎 https://x.com/appjsconf
💎 https://x.com/swmansion
Models designed to run on edge devices are more popular than ever, and running them on phones with React Native seems straightforward, but it introduces a unique set of challenges you won't see coming. This talk draws on a year and a half of building react-native-executorch and shipping real applications with on-device inference. We'll walk through the problems that surface only once you actually try: app sizes exploding, models crashing on mid-range devices, batteries draining during continuous inference, and the UX challenges of long model loads. We'll cover both LLM and computer vision use cases, from on-device chat to real-time object detection, and the very different constraints each brings to mobile. You'll leave knowing which mobile AI use cases are production-ready today and which ones still need more time.
Don't miss any updates about App.js Conf and follow us on X:
💎 https://x.com/appjsconf
💎 https://x.com/swmansion
