Google AI Edge Gallery brings local Gemma 4 to the Play Store

TL;DR
- Google’s AI Edge Gallery app is now officially available on the Google Play Store and Apple App Store.
- The app allows users to run the brand-new Gemma 4 model entirely on-device, requiring no internet connection.
- Users can now perform complex tasks, like coding and document summarization, with total offline privacy.
For the past year, running a high-end flagship AI model entirely on your device was a task largely reserved for developers and first adopters who wanted to try the next big thing. That changed this week. Google has officially launched its AI Edge Gallery to both Google Play and the App Store, allowing anyone with an Android device or iPhone a chance to run the all-new Gemma 4 mode entirely offline.
Until now, testing Google’s “Edge” AI required sideloading APKs and navigating complete setup screens. By moving AI Edge Gallery to the mainstream, Google is signaling that on-device AI is ready for prime time. It’s no longer a science experiment; it could quickly become an important AI tool for everyone.
In its current state, the app acts as a local sandbox for Google’s Gemini models. Unlike Gemini, which sends your data to Google for processing, AI Edge Gallery downloads the brain of AI directly to your device. In other words, you can now use AI anywhere without an internet connection, such as being in the middle of the ocean on a cruise ship or up at 35,000 feet on a plane. Better still, because the AI stays with you, you gain a certain level of privacy you didn’t have before.
The headline feature of the public release is support for Gemma 4. This isn’t just a minor iteration; Gemma 4 is built on the same architecture as Gemini 3, but offers improvements in logic, multilingual support, and a 256K context window.
What to expect from Gemma 4
Gemma 4 requires devices running at least Android 12 or iOS 17, and you can now perform tasks that used to require a data connection. For example, you can now utilize the Prompt Lab to summarize large PDFs or write complex code snippets locally. With Ask Image, it’s now possible to identify objects, plants, or texts without your photos without connecting first to Google’s servers. There’s also a new Audio Scribe feature for offline transcription. And then there’s perhaps the most important improvement: It’s now possible to have a fluid conversation that doesn’t suddenly disappear because of a poor internet connection.
Getting started with Gemma 4 is relatively pain-free. In your store, download AI Edge Gallery and the app will guide you through selecting a model. If you have a flagship device like the Pixel 10 Pro XL, look for the Gemma E4B model. It’s the “most intelligent” version and ideally suited for summarizing long documents, writing code, or complex travel planning. If you have a mid-range phone or just want a faster experience, download the E2B model. Regardless, because the app is now in the Play Store, it will dynamically switch between these two paths depending on your device’s battery life or thermal levels.
Finally, it should be noted that because AI Edge Gallery has moved into official stores, you no longer have to have a Hugging Face account, which means no more tokens and developer accounts to use it.
For now, I wouldn’t recommend AI Edge Gallery as a Gemini replacement. It can’t do everything that Gemini does, at least for now such as being able to check your emails or toggling on your flashlight. And yet, it’s nice to see Google opening up Gemma 4 for everyone if for no other reason than showing us all what’s coming soon.
Thank you for being part of our community. Read our Comment Policy before posting.



