Back to directory

Localight

A SwiftUI showcase chat app demonstrating Apple's on-device Foundation Models on iOS 26 and iOS 27, providing fast, private AI chat with no app server and no data leaving the device.

Localight

Localight is a native SwiftUI chat app that runs entirely on Apple's on-device Foundation Models on iOS 26 and iOS 27, showing how to integrate the Foundation Models framework into a native iOS app through SystemLanguageModel.default. Once the operating system has downloaded and prepared the model assets, response generation works fully offline with no app server involved.

Apple's third-generation model family includes two on-device models (a dense 3-billion-parameter model and a sparse multimodal 20-billion-parameter model) and three larger server-based models; the operating system chooses which one to use, and Localight does not select a specific model itself. The app keeps no conversation history and offers a minimalist chat interface.

The README states plainly that Localight is a demonstration app, not production-ready, and that model output may be inaccurate, incomplete, or misleading.

More in Inference

See all 770
  • OpenDecision

    deepanwadhwa.github.io

    An open-source Python package that answers typed questions about application state or documents using a local natural-language-inference model, returning structured choice, score, yes/no or evidence-relation values instead of free text.

  • Lifeboat

    iterate.ai

    A local inference server that puts an OpenAI- and Anthropic-compatible API in front of open-weight language models running on hardware you control, from a laptop to a GPU cluster.

  • Claude S40

    github.com

    An unofficial Claude chat client for 2007-era Nokia Series 40 phones, built as a Java ME app plus a small Go relay server, with web search, reading mode and saved chats.