Google AI Edge Gallery

2026-04-06 · Source: Simon Willison's Weblog · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Software Development & Engineering, Emerging Technologies & Innovation · Depth: Intermediate, quick

Summary

Google has released the "Google AI Edge Gallery" app, enabling users to run Gemma 4 (E2B and E4B sizes) and some Gemma 3 models directly on an iPhone. The E2B model, a 2.54GB download, demonstrates fast and useful performance. The app includes features like image-based questioning and audio transcription for up to 30 seconds using the smaller Gemma 4 models. It also showcases an "agent skills" demo, illustrating tool calling against eight interactive HTML widgets such as an interactive map, Wikipedia query, and QR code generator. This marks a notable instance of a local model vendor providing an official app for on-device model execution on iPhones, though it currently lacks permanent conversation logs.

Key takeaway

For AI Product Managers evaluating edge AI solutions, Google's AI Edge Gallery demonstrates a viable path for deploying large language models and agentic capabilities directly on mobile devices. Your team should explore the app's performance and tool-calling features to understand the potential for offline functionality and reduced latency in future applications, despite the current lack of persistent conversation logs.

Key insights

Google's new app enables on-device execution of Gemma models and tool calling directly on iPhones.

Principles

• On-device AI enhances privacy and speed.
• Tool calling expands local model utility.

In practice

• Run Gemma 4 E2B/E4B models locally.
• Experiment with agentic tool calling demos.

Topics

• Google AI Edge Gallery
• Gemma Models
• On-device AI
• iPhone Applications
• Tool Calling

Best for: NLP Engineer, AI Product Manager, Entrepreneur, AI Engineer, Machine Learning Engineer, AI Student

Related on AIssential

• A Smarter Google AI Edge Gallery: MCP integration, notifications, and session continuity — Google AI Edge Gallery enhances on-device AI with MCP, notifications, and persistent chat for agentic mobile experiences.
• How to Run Gemma 4 on Your Phone Without Internet: A Hands-On Guide — Gemma 4 enables private, offline AI capabilities directly on mobile devices via the Google AI Edge Gallery app.
• TLMs: Tiny LLMs and Agents on Edge Devices with LiteRT-LM — Cormac Brick, Google — Edge AI enables low-latency, private, and cost-effective on-device model deployment, leveraging tiny LLMs and agent skills.
• Apple's Edge Moat & the AI Frontier Gap — Apple's AI strategy requires both competitive edge inference and strong frontier models to succeed against cloud-based offerings.
• Accelerating on-device AI: A look at Arm and Google AI Edge optimization — Arm SME2 and Google AI Edge enable efficient, high-quality on-device generative AI inference on CPUs.
• Sarvam Edge: A Beginner’s Guide to On-Device AI for India — On-device AI offers instant, private, and offline functionality without per-query costs.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Simon Willison's Weblog.