Google upgrades its AI clothing try-on tool: Now a single selfie creates a full-body virtual model

Google

Google is expanding its AI-powered clothing try-on feature, allowing users to generate a full-body digital model of themselves using only a selfie. Previously, the tool required a full-body photo, but the update uses Google’s Nano Banana model (Gemini 2.5 Flash Image) to create a realistic, full-body avatar from a single face shot.

Once users upload a selfie, they can choose their clothing size and receive multiple generated full-body images. They can then select one image to serve as their default virtual try-on model. For those who prefer it, the option to upload a full-body picture remains available, alongside a set of diverse preset models.

The expanded try-on capability rolls out today in the United States.

Google introduced the original AI try-on feature in July, enabling users to preview apparel items across Search, Google Shopping and Google Images. Anyone browsing clothing can tap the “try it on” icon on a product listing to see how the item might look on them.

The update comes amid Google’s growing push into virtual fashion tools. The company has a standalone AI try-on app called Doppl, which visualises outfits using generative models. This week, Google revamped Doppl with a shoppable discovery feed that shows personalised outfit suggestions, AI-generated videos of real products and direct purchase links. Nearly everything in the feed is interactive, making virtual try-ons a more integrated shopping experience.

By combining selfie-based modelling, personalised style recommendations and shoppable AI content, Google is betting that virtual try-ons will become a standard part of online fashion browsing — mirroring the short-video formats users already engage with on Instagram and TikTok.