OpenAI introduced a virtual try-on feature for clothing and accessories within ChatGPT on both mobile and web platforms, allowing users to upload selfies or reference photos to see how specific items fit them. The update arrives as consumer adoption of artificial intelligence for online retail searches accelerates across the United States.
Virtual Fitting Tools Land in ChatGPT Search Results
When searching for clothes and accessories inside ChatGPT, users will now see a dedicated “Try on” button on product listings, according to an OpenAI note reported by pymnts.com. Selecting this button prompts the user to take or upload a selfie or a full-height photo to generate a virtual try-on image using ChatGPT images, with reference photos saved for future sessions. Alternatively, users can upload pictures of clothing items or screenshots from retail websites directly into a conversation to generate matching try-on visuals. TechCrunch reported that the feature utilizes the ChatGPT Images 2.5 model, which OpenAI states reproduces lighting and textures more accurately, follows editing instructions more precisely, and generates results faster.
New Library Controls and Visual Discovery Tools
Alongside the fitting room functionality, OpenAI added a Favorites feature that lets users save products and organize them into folders within their ChatGPT Library, where virtual fitting images are also stored. This capability builds upon visual shopping updates introduced in March, which included support for product discovery through the Agentic Commerce Protocol. These additions enable users to browse visually, upload photos to find similar items, and compare options side-by-side regarding prices, reviews, and features. According to a PYMNTS Intelligence report published in October titled “From Prompt to Purchase: Who Wins the AI Shopper?,” nearly 23% of consumers in the United States, representing almost 61 million people, initiate online retail searches using artificial intelligence, with clothing, shoes, and accessories accounting for about 28% of those recent purchases.
Platform Integration Across the Retail Sector
The integration by OpenAI reflects a broader industry push by platform companies, retailers, and startups to embed generative AI virtual try-on technology directly into product pages, search results, and checkout flows. Google expanded its own AI-powered virtual try-on technology to include shoes in October 2025, noting at the time that shoppers in the U.S. had shared their try-on images significantly more often than standard product listings.
Frequently Asked Questions About the ChatGPT Try-On Feature
- Where is the virtual try-on feature available?
- The capability is accessible in ChatGPT on both mobile and web platforms, appearing directly on product listings for clothes and accessories.
- How does the AI generate the virtual fitting image?
- Users click the “Try on” button in search results and upload a selfie or full-height photo, or they can upload a picture of a garment and ask the assistant to generate the image using the ChatGPT Images 2.5 model.
- What percentage of U.S. consumers start online retail searches with AI?
- According to an October PYMNTS Intelligence report, nearly 23% of consumers in the United States, or almost 61 million people, start their online retail search with artificial intelligence.
More on United States
Related reading