ChatGPT shopping adds virtual Try On and Favorites
On October 1, 2026 OpenAI gave ChatGPT's shopping experience two new features. With Try On, a user uploads a selfie or a full-body photo and sees how clothing and accessories would look on them; Favorites keeps products they like in the app's Library. The images are generated by the new ChatGPT Images 2.5 model.
A try-on button lands in shopping results
Writing for TechCrunch on October 1, 2026, Sarah Perez reported that OpenAI had announced two shopping additions for ChatGPT: virtual try-on for clothing and accessories, and Favorites, a way to keep products for a later visit. TechCrunch describes the rollout as global. Engadget covered the news the same day, and Search Engine Roundtable relayed short passages written by OpenAI.
In Perez's words, "This option will appear as a new Try On button in ChatGPT's shopping results." The user supplies a selfie or a full-body photo and gets back an image of the item on their own body. According to Search Engine Roundtable, the feature is rolling out on mobile and web.
Screenshots work too, and the Library has folders
Try-on is not limited to products ChatGPT recommends. TechCrunch says a user can upload a picture of an item, a web screenshot for example, and ask for a try-on, while Engadget notes that, in theory, a screenshot of any piece of clothing could be used, including items the chatbot never surfaced. The images come from OpenAI's new ChatGPT Images 2.5 model. As relayed by TechCrunch, the company says the model renders lighting more naturally and textures more richly, sticks to editing instructions more reliably and cuts generation latency.
Favorites stores discovered products in the app's Library. According to Search Engine Roundtable, saved products can be sorted into folders, and the reference photo uploaded for try-on is kept for future sessions and can be changed or deleted under Settings > Personalization. TechCrunch adds two more use cases: describing a style and asking ChatGPT to shop for the pieces that complete it, or uploading photos of celebrity outfits and asking it to track down items that can be bought.
The fitting-room moment moves ahead of the store
Product discovery, and the question of how something will look on you, increasingly gets answered inside a chat interface before the shopper ever reaches a store's website. That raises the stakes on product images and product data being clean and readable by AI systems. And because a shopper can try on an item from any store via a screenshot, small clothing brands can end up inside this flow indirectly.
Context matters here. Engadget points out that Google brought the same idea to Search in 2025, and an update to its story records that OpenAI's Instant Checkout feature has been discontinued. Engadget also flags a privacy point: images uploaded from a personal account may be used for model training by default unless the user opts out. For a feature built on selfies and full-body photos, that detail deserves attention.
The sources say nothing about Türkiye, so this section is commentary
None of the three sources mentions Türkiye. TechCrunch calls the launch global but gives no country list. What follows is therefore not drawn from the sources but is explicitly UNALSOFT commentary. To list the unknowns one by one: the sources do not say whether the feature is available to users in Türkiye today, whether it behaves the same in a Turkish-language interface, which subscription plans include it, or which retailers' products appear in the shopping results.
Our reading: for a business in Türkiye selling clothing and accessories, the first question is how its products show up on AI shopping surfaces like this one. Since try-on also works from a screenshot, a shopper may judge an item purely from your product photo, so a sharp, single-product, colour-accurate image now carries more weight. We looked at how product data is read on AI surfaces from Google's side on September 17; that article is independent of this news.
UNALSOFT's take
We read this less as a new gadget and more as the shop window changing places. If a shopper sees a product inside a chat and tries it on there, the thing that decides the moment is no longer the store page design but the quality of the product image and product information. The practical consequence is plain: clean single-product images, consistent names and descriptions, and up-to-date stock and colour details for every item. That is why on the e-commerce panel side we treat keeping the catalogue readable for both people and AI tools as its own piece of work. Because the sources carry no data about Türkiye, this article makes no claim about local availability.
How do your products look in the AI shop window?
A short conversation is enough to get your product images and catalogue data into shape for both customers and AI tools.