google-ai-edge/gallery

A gallery that showcases on-device ML/GenAI use cases and allows people to try and use models locally.

What it solves

It provides a centralized platform for running powerful open-source Large Language Models (LLMs) directly on mobile devices and macOS. This eliminates the need to send data to external servers, ensuring total privacy, offline functionality, and fast response times.

How it works

The application leverages Google AI Edge and the LiteRT lightweight runtime to execute optimized models on local hardware. It integrates with Hugging Face for model discovery and downloading, allowing users to load official models (like the Gemma 4 family) or their own custom models.

Who it’s for

It is designed for developers and AI enthusiasts who want to experience, test, and evaluate the performance of on-device generative AI models on their specific hardware.

Highlights

  • Agent Skills: Augments LLMs with tools like Wikipedia, interactive maps, and visual summary cards.
  • AI Chat with Thinking Mode: Supports multi-turn conversations and a reasoning mode to visualize the model's step-by-step logic.
  • Multimodal Capabilities: Includes "Ask Image" for visual analysis and "Audio Scribe" for real-time voice transcription and translation.
  • Prompt Lab: A workspace for testing prompts with granular control over parameters like temperature and top-k.
  • Model Benchmarking: Tools to manage a model library and run performance tests on local hardware.
  • Mobile Actions: Uses a fine-tuned FunctionGemma 270m model to power offline device controls and automated tasks.

Related

  • Project
    dineshsoudagar/local-llms-on-android

    An Android application that enables fully offline, private LLM chat with support for voice, image, and camera input using ONNX and LiteRT backends.

  • Dispatch
    LLM Inference on Edge: Running Local LLMs via React Native

    Hugging Face provides a comprehensive guide to building a privacy-focused mobile application using React Native and llama.rn to run quantized GGUF models locally on Android and iOS.

  • Project
    scaleapi/llm-engine

    An open-source engine for fine-tuning and serving large language models, providing tools to deploy models on hosted or self-managed Kubernetes infrastructure.

  • Project
    google-ai-edge/LiteRT-LM

    A production-ready orchestration layer for running high-performance LLMs on-device across mobile, web, and desktop platforms using LiteRT.

  • Project
    Sentient-OS-Labs/sentient-os

    An on-device AI operating system for macOS that builds a private, local knowledge base of your digital life to power proactive agents and computer-use automation.