Google will allow app developers to use the Gemini Nano model on devices.

Google to Allow App Developers to Use Gemini Nano Model on Devices

Google has quietly released new artificial intelligence feature documentation for developers. The update to the ML Kit SDK will add API support for on-device generative AI features via Gemini Nano. It is built on AI Core, similar to the experimental Edge AI SDK, but it can plug into existing models and comes with a set of predefined features that developers should be able to implement easily. The GenAI API of ML Kit will enable applications to perform summarization, proofreading, rewriting, and image description without sending data to the cloud. The output quality may also vary depending on the version of Gemini Nano on the phone. The standard version (Gemini Nano XS) is about 100MB in size, but the Gemini Nano XXS on the Pixel 9a is only a quarter of that size. It only supports text and has a much smaller context window.

—— Android Authority, Ars Technica

via Wind Vane Reference Express - Telegram Channel