ChatGPT can now virtually try on clothes for you

OpenAI is expanding ChatGPT's shopping features with a virtual try-on tool that lets users see clothing and accessories overlaid onto photos of themselves. Rather than browsing static product images, shoppers can now get a more personalized sense of how an item might look before committing to a purchase. The feature appears to be part of a broader effort by OpenAI to make ChatGPT a more practical, everyday utility beyond text and image generation.

The virtual try-on capability leans on generative image models to composite selected products onto user-provided photos. This kind of technology has existed in various forms across retail apps and dedicated fashion platforms for several years, but integrating it directly into a general-purpose AI assistant gives it a wider potential audience. Users who are already turning to ChatGPT for product recommendations now have a way to visualize those recommendations without leaving the interface.

Alongside the try-on feature, OpenAI is introducing a Favorites library - a way for users to bookmark products surfaced during a conversation and return to them later. This adds a layer of continuity to the shopping experience, addressing one of the more practical shortcomings of chat-based discovery, where items found in one session could easily be lost. Together, the two features push ChatGPT closer to the kind of end-to-end shopping experience traditionally associated with dedicated e-commerce platforms.

The timing reflects increasing competition among AI assistants to become the starting point for online shopping. Google has integrated shopping features across its AI products, and a number of startups have focused specifically on AI-driven fashion discovery. For OpenAI, adding these capabilities to ChatGPT is a way to deepen user engagement and demonstrate practical value in a domain - retail and fashion - where personalization has long been a core challenge.