OpenAI Adds Virtual Try-On Features to ChatGPT
OpenAI has launched virtual try-on and favoriting features for ChatGPT, leveraging its new image model to turn the AI assistant into a highly visual personal shopping companion.

OpenAI has introduced two new shopping capabilities globally for its conversational assistant, ChatGPT. Users can now virtually try on clothing and accessories using their own photos, as well as save items they like to a dedicated favorites library. This rollout marks the company's latest attempt to capture a share of the e-commerce market, following a previously discontinued instant checkout feature that failed to gain traction.
The new features are powered by OpenAI's recently released ChatGPT Images 2.5 model. According to the company, this updated model delivers more realistic lighting, richer textures, and faster image generation times. It is also designed to follow editing instructions more reliably, which is crucial for accurately superimposing garments onto user-provided photos.
To use the virtual try-on tool, shoppers can upload a selfie or a full-body image to ChatGPT. A new 'Try On' button will then appear alongside shopping search results, allowing the AI to visualize how the clothing would look on the user. Alternatively, users can upload a screenshot of an item from another website and ask the assistant to generate an image of them wearing it. The accompanying Favorites feature saves these generated try-on images and discovered products to an in-app Library for future reference.
Beyond trying on clothes, ChatGPT can now assist with broader fashion discovery. Users can describe a specific aesthetic to receive shopping recommendations, or upload photos of celebrities to find matching, purchasable items online. This expansion positions OpenAI to compete directly with Google, which launched its own virtual try-on tool last year, and Pinterest, both of which have long dominated the visual search and fashion inspiration space.
This is our own summary of reporting by TechCrunch AI



