// AI NEWS · FEATURE

ChatGPT Can Now Virtually Try On Clothes Inside Its Shopping Results

OpenAI began a global rollout of Virtual Try On and a Favorites library on October 1, 2026. Upload a photo, tap Try On in a shopping result, and ChatGPT generates an image of the item on you — powered by the new Images 2.5 model.

2026-10-01 · by Moe Ameen

What happened

OpenAI announced on October 1, 2026 that it is rolling out two new shopping features in ChatGPT globally: Virtual Try On and Favorites. With Virtual Try On, you upload a selfie or a full-body photo and tap a new Try On button that appears on shopping results, and ChatGPT generates an image showing how an article of clothing or an accessory might look on you. You can also upload an image of an item yourself — for example a screenshot from the web — and ask ChatGPT to try it on.

The feature runs on ChatGPT Images 2.5, OpenAI's image model released in September 2026, which the company describes as producing more natural lighting and richer textures, following editing instructions more reliably, and generating faster than the prior version. That model quality is what makes a try-on visualization plausible rather than cartoonish.

The companion feature, Favorites, lets you save products you find into a Library inside the app so you can return to them later; your try-on images are saved alongside the items. Both arrive as OpenAI keeps experimenting with how ChatGPT can help people shop: the company quietly shut down its earlier Instant Checkout in-chat buying feature in March 2026 after it reportedly underperformed, and has since leaned on discovery features like product results instead.

OpenAI is clear that the generated images are visualizations, not a promise: a try-on image does not guarantee fit or size. Treat availability, exact placement of the Try On button, and which plans see it first as a snapshot of a rollout in progress, and confirm specifics against OpenAI's own help pages, since shopping features change quickly.

Why it matters for creators

  • Discovery is moving into the assistant. People increasingly start a product search in ChatGPT, and a try-on button inside that result makes the chat a storefront, not just a referrer — another surface brands have to show up in.
  • It raises the bar on product visuals. If shoppers can picture an item on themselves instantly, flat catalog photos feel thinner; the brands that win attention will be the ones producing richer on-model, lifestyle, and demo content.
  • It is a buyer-side tool, not a seller-side one. The try-on happens inside OpenAI's app for the shopper. You cannot publish from it, brand it, or point it at your own catalog — it does nothing to produce the content that gets you discovered in the first place.
  • Fit is explicitly not guaranteed. The image is a styling preview, so returns, sizing guidance, and honest product detail still matter — and still have to be communicated in content you control.
  • The same Images 2.5 model powers both this and serious content workflows. The quality jump that makes try-on convincing is the same jump that makes AI-generated product and scene imagery good enough to publish.

How to act on this with Kompozy

Virtual Try On is a consumer feature that lives inside ChatGPT: it helps a shopper picture an item on themselves, and that is where it ends. For a brand or creator there is nothing to publish from it and no way to point it at your catalog — it is a reason to be found, not a tool that gets you found. The job it quietly raises the bar on is the one it does not touch: producing enough strong, on-brand product content that you actually surface when someone shops, and keeping it flowing across every platform where buyers look.

That production-and-distribution job is what [Kompozy](/) is built for. Feed it a product, a drop, or a single campaign idea and it generates the full shoppable set — Photo Posts and Infographic Photo scene imagery on the same gpt-image family that powers Images 2.5, Persona Photos with a face-locked model, [Carousel Posts](/glossary/hyperframes) that show a product from several angles, Clipped Shorts and Listicle Video for Reels, TikTok, and Shorts, plus a Blog Article and an Email Newsletter to make the launch searchable and send it to your list. A [Persona Brief](/glossary/persona-brief) keeps every piece in one voice, and autopilot with a per-post review step schedules and publishes the lot across eight social platforms plus blog and email. ChatGPT lets a shopper try your product on; Kompozy produces and ships the content that put it in front of them.

Quick takeaways

  • OpenAI started rolling out Virtual Try On and Favorites in ChatGPT globally on October 1, 2026.
  • Upload a selfie or full-body photo, tap Try On in a shopping result, and ChatGPT renders the item on you; you can also upload an item image yourself. It runs on ChatGPT Images 2.5.
  • OpenAI says the images are visualizations and do not guarantee fit or size. Favorites saves products — and your try-on shots — to a Library.
  • It is a buyer-side feature inside ChatGPT; you cannot publish or brand from it. Kompozy generates the shoppable product content — scenes, carousels, shorts, blogs, newsletters — and ships it across eight platforms plus blog and email.

Frequently asked questions

What is ChatGPT Virtual Try On?

It is a shopping feature OpenAI began rolling out globally on October 1, 2026. You upload a selfie or full-body photo and tap a Try On button on a shopping result, and ChatGPT generates an image showing how a clothing item or accessory might look on you. You can also upload an item image, such as a web screenshot, and ask ChatGPT to try it on. It runs on the ChatGPT Images 2.5 model.

Is the ChatGPT try-on image accurate for fit and size?

No. OpenAI is explicit that the generated images are visualizations and do not guarantee fit or size. Treat a try-on as a styling preview that helps you picture an item, not as a replacement for a size chart, measurements, or a brand's own fit guidance.

What is the Favorites feature in ChatGPT?

Favorites, launched alongside Virtual Try On, lets you save products you discover into a Library inside the ChatGPT app so you can come back to them later. Your try-on images are saved alongside the saved items.

Can a brand or creator publish content from ChatGPT Virtual Try On?

No. Virtual Try On is a buyer-side tool that produces a private visualization inside the shopper's chat — there is nothing to brand or publish, and it does not point at your catalog. To produce and distribute the shoppable content that gets you discovered, you need a generation and publishing engine. Kompozy turns a product into scene images, carousels, shorts, a blog, and a newsletter, then schedules and publishes them across eight social platforms plus blog and email.

Related news

← All AI news · Get started →