Does Amazon Have Virtual Try-On for Clothes? What It Covers and What It Does Not

A shopper asking whether Amazon has virtual try-on for clothes usually means something specific: can I see a dress or a jacket on my own body before buying it. The honest answer is mixed. Amazon has real, working virtual try-on technology, and it has expanded past shoes since it first launched in 2022, but the categories it actually covers today are narrower than the underlying technology supports.

Does Amazon have virtual try-on for clothes?

Amazon's virtual try-on lets a shopper visualize an item on their own body through the Amazon Shopping app, using AR and AI to render the product onto an uploaded photo or a matched avatar. It started with shoes in 2022, later added eyewear, and has since added a limited set of t-shirts. That is the current scope: shoes, eyewear, select t-shirts.

What clothing categories does Amazon's virtual try-on actually support?

• Shoes: the original category, launched in 2022, covering thousands of styles from major brands using AR to place a shoe on a photo of the shopper's foot.
• Eyewear: select sunglasses and glasses styles, shown on the shopper's own face.
• Select t-shirts: a limited set of t-shirt styles, shown on a size-matched 3D avatar rather than the shopper's own uploaded photo in every case.
• Not yet covered: full clothing categories including dresses, pants, jackets, and suits, according to AWS's own developer documentation for the underlying Nova Canvas model.

That gap between what the underlying AI model can technically do and what actually ships to shoppers points to a product and rollout decision at the marketplace level, not a hard technical ceiling.

What technology is behind Amazon's virtual try-on?

Amazon's virtual try-on runs on Amazon Nova Canvas, a generative AI image model built by AWS. Nova Canvas's virtual try-on capability takes two inputs, a photo of the person and a product reference image, and uses a masking system to decide which part of the photo to edit. AWS's own documentation lists masking modes for upper body, lower body, full body, and footwear regions specifically, meaning the model architecture already supports full-garment editing, not just shoes and t-shirts.

1. A shopper or developer supplies a source photo and a product image.
2. Nova Canvas identifies the relevant region of the photo using the selected mask mode.
3. The model generates a new image with the product rendered into that region, keeping the rest of the photo intact.
4. Up to five image variations can be generated per request, with one returned by default.

What happened to Amazon's Try Before You Buy program?

Amazon ended its physical Try Before You Buy program, which let a customer order several items, try them on at home, and return whatever did not fit or work, in January 2025. Ending a program that let people physically handle and wear a garment before paying, then leaning more heavily on AI virtual try-on and AI powered size recommendations, is a real signal about where Amazon expects fit confidence to come from going forward.

What does this mean if you sell clothing outside Amazon?

A marketplace-level try-on feature works well where it exists, but it is scoped to whatever the marketplace decides to build, on the categories it chooses, inside its own app. That is a different situation from earlier coverage on this blog of a brand building its own in-house try-on tool, or a search engine surfacing try-on inside its own results.

• Category coverage is decided by the platform, not the seller. A retailer selling dresses, jackets, or suits on Amazon cannot add virtual try-on to those listings today, since the consumer feature does not extend that far.
• The experience lives inside one app. A shopper needs the Amazon Shopping app specifically, not a browser or a store's own product page.
• A merchant's own store is not bound by the same rollout schedule. Virtual try-on installed directly on a Shopify or WooCommerce product page covers whatever categories the merchant actually stocks, including the full clothing range Amazon's consumer app has not yet extended to.

For how this compares to a search engine adding try-on to its own results, see virtual try-on in AI shopping search. For how individual fashion brands are building their own tools rather than relying on a platform, see big fashion brands using AI virtual try-on.

Corlen puts virtual try-on directly on a merchant's own product pages, covering the full range of garments a store actually sells, on Shopify and WooCommerce, an in-store kiosk, or through a developer API for a custom build, without waiting on which categories any single marketplace decides to support next.

See it on your own photo: try Corlen for free, no account needed for a first result.