Back to the chart

google-ai-edge/gallery

On-device generative AI model showcase and benchmark app

Chart
Not on a chart right now
Stars
25.4k

Open on GitHub

What it is

A mobile app for running and testing open-source LLMs and multimodal AI models locally without an internet connection. Built on Google AI Edge and LiteRT, it lets users try on-device capabilities ranging from prompt experiments to speech and media search.

Who it is for

Developers and researchers who want to evaluate on-device AI models and use cases directly on mobile hardware.

Why it is trending

It gained 100 stars today, reaching 25,372 total stars, following its 1.0.23 release on November 4, 2026. It stands out by running all inference 100% offline on device hardware with built-in benchmarking.

Main features

  • Modular Agent Skills for tool integration
  • AI chat with on-device speech recognition and text-to-speech
  • Offline semantic media and video moment search with EmbeddingGemma 2
  • Custom model loading and per-device hardware benchmarks

Summarized on November 19, 2026 from the README and repository data. It can be wrong; check the source before relying on it. Report a wrong summary