Google AI Edge Gallery

Simon Willison ·

Google releases AI Edge Gallery app for iOS, enabling local Gemma 4 inference on-device with image Q&A and audio transcription capabilities.

Categories: Products to Try

Excerpt

<p><strong><a href="https://apps.apple.com/nl/app/google-ai-edge-gallery/id6749645337">Google AI Edge Gallery</a></strong></p> Terrible name, really great app: this is Google's official app for running their Gemma 4 models (the E2B and E4B sizes, plus some members of the Gemma 3 family) directly on your iPhone.</p> <p>It works <em>really</em> well. The E2B model is a 2.54GB download and is both fast and genuinely useful.</p> <p>The app also provides "ask questions about images" and audio transcription (up to 30s) with the two small Gemma 4 models, and has an interesting "skills" demo which demonstrates tool calling against eight different interactive widgets, each implemented as an HTML page (though sadly the source code is not visible): interactive-map, kitchen-adventure, calculate-hash, text-spinner, mood-tracker, mnemonic-password, query-wikipedia, and qr-code.</p> <p><img alt="Screenshot of an &quot;Agent Skills&quot; chat interface using the Gemma-4-E2B-it model. The user prompt reads &quot;Show me the Castro Theatre on a map.&quot; The model response, labeled &quot;Model on GPU,&quot; shows it &quot;Called JS skill 'interactive-map/index.html'&quot; and displays an embedded Google Map centered on a red pin at The Castro Theatre in San Francisco, with nearby landmarks visible including Starbelly, Cliff's Variety, Blind Butcher, GLBT Historical Society Museum, and Fable. An &quot;Open in Maps&quot; link and &quot;View in full screen&quot; button are shown. Below the map, the model states &quot;The interactive map view for the Castro Theatre has been shown.&quot; with a response time of 2.4 s. A text input field with &quot;Type prompt...&quot; placeholder, a &quot;+&quot; button, and a &quot;Skills&quot; button appear at the bottom." src="https://static.simonwillison.net/static/2026/gemini-agent-skills.jpg" /></p> <p>(That demo did freeze the app when I tried to add a follow-up prompt though.)</p> <p>This is the first time I've seen a local model vendor release a