Posted by Dave Burke, VP of Engineering
Final week we unveiled our most succesful basis mannequin, Gemini. Gemini is multimodal – it may well settle for each textual content and picture inputs. We launched a approach for Android builders to leverage our smallest mannequin Gemini Nano, on-device. That is obtainable on choose gadgets by AICore, a system service that handles mannequin administration, runtimes, security options and extra, simplifying the work for builders. And at present, we’re introducing new methods for Android builders to entry the Gemini Professional mannequin – which runs off-device, in Google’s knowledge facilities.
App improvement with Gemini Professional
Gemini Professional is accessible by way of the Gemini API, and it’s our greatest mannequin for scaling throughout a variety of textual content and picture reasoning duties. To simplify integrating Gemini Professional, you should use the Google AI SDK, a shopper SDK for Android. This SDK permits direct integration from Android apps and removes the necessity for builders to construct and handle their very own backend infrastructure, decreasing improvement prices and enhancing velocity.
Google AI Studio supplies a streamlined approach for builders to combine the Gemini Professional mannequin, craft prompts, create API keys, and effortlessly rework concepts into AI apps. After getting developed your immediate in Google AI Studio, you’ll be able to merely click on on the “Get code” motion to generate a Kotlin code snippet, and begin integrating Gemini at present utilizing the Google AI SDK for Android.
We’re additionally making it simpler for builders to make use of the Gemini API immediately within the latest preview version of Android Studio. We’re introducing a new project template for builders to get began with the Google AI SDK for Android immediately. You’ll profit from Android Studio’s enhanced code completion and lint checkers, serving to with API keys and safety.
To leverage the brand new template in Android Studio, begin a brand new challenge by File > New > New Mission and decide the Gemini API starter template. This template supplies a pre-configured challenge with the required code to make use of the Gemini API. After selecting a challenge identify and placement, you can be prompted to generate an API key in Google AI Studio, and requested to enter it in Android Studio. Android Studio will mechanically arrange the challenge for you with the Gemini API connection, simplifying your workflow.
Alternatively, you’ll be able to import the generative AI code sample and set it up in Android Studio by File > New > Import Pattern, and trying to find “Generative AI Pattern”.
Get began constructing AI-powered options and Android apps utilizing Gemini Pro.