Posted by Paris Hsu – Product Supervisor, Android Studio
At each stage of the event lifecycle, Gemini in Android Studio has turn into your AI-powered companion, making it simpler to construct top quality apps. We’re excited to announce a big growth: Gemini in Android Studio now helps multimodal inputs, which helps you to connect photos on to your prompts! This unlocks a wealth of latest potentialities that enhance workforce collaboration and UI improvement workflows.
You’ll be able to check out this new function by downloading the most recent Android Studio canary. We’ve outlined a couple of use instances to attempt, however we’d love to listen to what you suppose as we work by bringing this function into future secure releases. Test it out:
Picture attachment – a brand new dimension of interplay
We first previewed Gemini’s multimodal capabilities at Google I/O 2024. This expertise permits Gemini in Android Studio to know easy wireframes, and rework them into working Jetpack Compose code.
You will now discover a picture attachment icon within the Gemini chat window. Merely connect JPEG or PNG information to your prompts and watch Gemini perceive and reply to visible data. We have noticed that photos with robust shade contrasts yield the perfect outcomes.


We encourage you to experiment with varied prompts and pictures. Listed here are a couple of compelling use instances to get you began:
- Fast UI prototyping and iteration: Convert a easy wireframe or high-fidelity mock of your app’s UI into working code.
- Diagram clarification and documentation: Acquire deeper insights into advanced structure or information movement diagrams by having Gemini clarify their elements and relationships.
- UI troubleshooting: Seize screenshots of UI bugs and ask Gemini for options.
Fast UI prototyping and iteration
Gemini’s multimodal help enables you to convert visible designs into purposeful UI code. Merely add your picture and use a transparent immediate. It really works whether or not you are working from your individual sketches or from a designer mockup.
Right here’s an instance immediate: “For this picture supplied, write Android Jetpack Compose code to make a display screen that is as near this picture as potential. Be certain to incorporate imports, use Material3, and doc the code.” After which you’ll be able to append any particular or extra directions associated to the picture.


For extra advanced UIs, refine your prompts to seize particular performance. As an example, when changing a calculator mockup, including “make the interactions and calculations work as you’d count on” leads to a totally purposeful calculator:


Observe: this function gives an preliminary design scaffold. It’s a superb “first draft” and your edits and changes might be wanted. Widespread refinements embrace guaranteeing right drawable imports and importing icons. Think about the generated code a extremely environment friendly start line, accelerating your UI improvement workflow.
Diagram clarification and documentation
With Gemini’s multimodal capabilities, it’s also possible to attempt importing a picture of your diagram and ask for explanations or documentation.
Instance immediate: Add the Now in Android structure diagram and say “Clarify the elements and information movement on this diagram” or “Write documentation about this diagram”.

UI troubleshooting
Leverage Gemini’s visible evaluation to establish and resolve bugs rapidly. Add a screenshot of the problematic UI, and Gemini will analyze the picture and counsel potential options. You can even embrace related code snippets for extra exact help.
Within the instance beneath, we used Compose UI verify and located that the button is stretched too vast in pill screens, so we took a screenshot and requested Gemini for options – it was in a position to leverage the window dimension lessons to offer the appropriate repair.

Obtain Android Studio right now
Obtain the most recent Android Studio canary right now to attempt the brand new multimodal options!
As all the time, Google is dedicated to the accountable use of AI. Android Studio will not ship any of your supply code to servers with out your consent. You’ll be able to learn extra on Gemini in Android Studio’s dedication to privateness.
We recognize any suggestions on stuff you like or options you want to see. If you happen to discover a bug, please report the difficulty and in addition try recognized points. Keep in mind to additionally observe us on X, Medium, or YouTube for extra Android improvement updates!