A Google AI model called Gemini Nano runs on-device and performs tasks such as reading and summarizing text, describing images, and more without requiring a network connection. Your device processes everything, so your data remains private. Now, Google is letting all Android app creators add Gemini Nano to their apps to provide useful features.
App makers can now test Gemini Nano using a new toolkit called the AI Edge SDK. This toolkit was previously only available to select app creators but now any developer can experiment adding AI to their Pixel phone apps. Support for more phones will come later.
The AI Edge SDK lets developers play with Gemini Nano’s abilities in text, like rewriting sentences more clearly, automatically responding to messages, checking for mistakes, or shortening long passages. In the future, the toolkit will support other types of media too, like photos. Even at this starting point, app creators can add capabilities like proofreading emails, summarizing articles, or generating smart replies in chat apps.
Developers get to customize how their apps use Gemini Nano’s powers. The toolkit helps control what information the AI analyzes and allows tuning the responses to best suit each application.
