Google has enhanced its AI-powered virtual try-on feature, allowing users to generate a full-body avatar from a simple selfie. The update eliminates the need for full-body photos and supports sizes from XS to 4XL+. This tool is now available in the US, building on the original version launched in May.
Google's latest improvement to its virtual try-on technology simplifies online shopping by using artificial intelligence to create personalized clothing visualizations. Previously, the feature required a full-body photograph to simulate outfits. The new version, revealed on Thursday, enables users to upload a selfie—even if it shows only the head and shoulders or a seated pose—and generate a complete digital figure.
Users select their preferred body size, ranging from XS to 4XL+, and the system produces a full-body image. This avatar is powered by Nano Banana, which is Google's Gemini 2.5 Flash Image model. Once created, shoppers can explore various garments on the virtual model, helping them visualize fits without physical try-ons.
The upgrade expands accessibility for the tool, originally unveiled in May during Google's I/O developers conference. That initial release already allowed virtual clothing trials, but demanded a standing, full-length photo. The refined approach accommodates more casual images, making it easier for people in the United States to experiment with styles like formal gowns or leather pants.
This development reflects ongoing efforts to integrate AI into e-commerce, reducing barriers to informed purchasing decisions. While the feature remains US-only for now, it maintains the option for full-body photo uploads for those who prefer it. By streamlining the process, Google aims to enhance user convenience in digital retail experiences.
