Forward Future Tools Library
Google AI Edge Gallery
Google AI Edge Gallery runs open-source generative AI models locally on phones and computers, giving developers and AI enthusiasts a way to test private, offline inference.
Try Google AI Edge Gallery →
github.com·Free
›What is Google AI Edge Gallery?
Google AI Edge Gallery is an open-source app for running on-device machine learning and generative AI models. It lets users download and try models such as Gemma locally, without sending prompts or data to a cloud server. The app also includes agent skills, image analysis, and audio features.
›What are the pros and cons of Google AI Edge Gallery?
Strengths
Runs models fully offline, so prompts and data do not need to leave the device
Free and open source, with no subscription or paywall identified
Supports multiple models, including mobile-optimized Gemma variants
Offers agent skills that extend models beyond basic chat
Works across iOS, Android, macOS, and APK installation
Trade-offs
Model downloads can require substantial device storage, with Gemma 4 E2B reported at 2.5GB
Response speed depends heavily on the device, and reviewers reported very long waits on a Fairphone 6
Reviewers found limited memory for ongoing context
Its system integration is limited compared with tools designed for daily assistant use
The experience is positioned more as an experimentation showcase than a finished everyday assistant
›What are Google AI Edge Gallery’s key features?
Runs open-source LLMs locally on mobile devices
Supports Gemma models optimized for on-device inference
Works offline without sending data to a server
Provides agent skills with tools such as Wikipedia, interactive maps, and visual summary cards
Supports image analysis and audio functions
Available through Google Play, the App Store, macOS, and an APK release
›What are the best use cases for Google AI Edge Gallery?
Testing local language models on a phone without cloud access
Evaluating privacy-sensitive AI interactions that keep data on the device
Experimenting with agent skills such as fact-finding and map-based interactions
Prototyping on-device image and audio AI experiences
Comparing Gemma and other mobile-optimized models on different hardware
›What is the pricing for Google AI Edge Gallery?
Free
›Who is Google AI Edge Gallery best for?
developersUseful for testing on-device inference, model behavior, and agent skills across mobile hardware.
AI researchersA practical way to evaluate local Gemma models and offline generative AI features.
soloWorks well for individuals who want private, offline experimentation without a subscription.
IoT engineersRelevant for exploring local AI patterns, though the app itself is focused on phones and computers.
Not for
- People who need consistently fast responses on lower-powered phones
- Users seeking a polished daily assistant with deep operating-system integration
- Teams that need shared cloud workspaces, centralized administration, or hosted model access
›What are the best Google AI Edge Gallery alternatives?
LM StudioChoose this instead if you want a local AI toolkit centered on downloading and running models from a desktop environment.→
OllamaChoose this instead if you want to run local language models such as Llama, DeepSeek-R1, Phi-4, Mistral, and Gemma through a developer-focused tool.→
LongCat Flash→
Google Gemini→