What is virtual try-on?
Virtual try-on is technology that shows how a garment, shoe or accessory would look on a person without it being physically worn, by rendering the item onto a photo, a live camera image or an avatar. The term covers two different jobs: shopper-facing try-on, where a customer sees the item on themselves, and brand-side catalog rendering, where a brand puts its garments on a model to produce product images.
Mixing up the two is the most common source of confusion when comparing tools. They use related technology, but they are bought by different teams, judged by different results and produce different outputs.
Shopper try-on and catalog rendering
Shopper-facing try-on sits in the store or an app. The shopper uploads a photo, turns on the camera or picks a body type, and sees the item on that. The person in the image is the customer. The image is usually seen once, by one person, and never becomes part of the catalog. It is judged on engagement, conversion and how believable the result looks to that shopper.
Brand-side catalog rendering sits in the content workflow. A content or studio team gives the tool a garment photo and a model, and gets back an image that is published on the product page, in a lookbook or in a campaign. The person in the image is a model, every visitor sees the image, and it has to hold up at full size next to the rest of the catalog. It is judged on garment fidelity, model consistency and resolution.
A good tool for one job is not automatically good for the other. A try-on widget does not produce catalog files, and a catalog tool does not put the shopper's own body in the frame.
How shopper try-on works
The earliest versions were augmented reality filters for glasses, makeup and shoes, which track the face or feet in a live camera feed and overlay a 3D model of the product. For clothing, current systems are mostly image-based: the shopper's photo and the product image go into a generative model that renders the garment onto the person. Some systems build an avatar from body measurements instead.
The limits are worth knowing. A rendered image shows how a garment might look, not how a specific size will fit. Most systems cannot know how a fabric stretches on a particular body, so try-on works best alongside size guidance rather than as a replacement for it.
How catalog rendering works
In catalog rendering the brand controls the inputs: a clean garment photo, such as a flat lay, packshot or ghost mannequin image, and a model, either generated or taken from an uploaded photo. The output replaces or supplements the on-model images from a photoshoot. Because these images represent the product to every shopper, the bar is different: colors, prints, logos and construction details must match the real garment, and the same model should look the same across a collection.
Cost and time
Shopper try-on is a store feature. It usually involves an integration into the shop and ongoing fees, and its value shows up in store metrics. Catalog rendering replaces part of the cost of a shoot: the garment still has to be photographed, but studio, crew and model days shrink. In both cases the hidden cost is quality control, because someone has to check that the result is believable and correct.
Virtual try-on and NotShot
For a comparison of tools in both groups, and where NotShot fits, read Best AI virtual try-on tools for fashion brands in 2026. Every render is judged against your upload. If a render fails the check, you get your credits back.
Related terms
- AI fashion model: a person in a fashion image created by AI
- On-model photography: the garment worn by a person
- AI product photography: product images created or edited with AI
- Garment fidelity: the main quality bar for catalog rendering
- Catalog photography: the standard image set for every product
- Return rate: what try-on is often meant to reduce
- Hallucination (AI images): invented or changed details
- Best AI virtual try-on tools for fashion brands in 2026
Questions we get asked
- What is the difference between virtual try-on and virtual fitting?
- The terms are often used interchangeably. Strictly, virtual fitting focuses on size and fit, often based on body measurements, while virtual try-on focuses on how the item looks on a person.
- Does virtual try-on reduce returns?
- It can help shoppers judge style and color before buying, which addresses some reasons for returns. A rendered image does not show how a specific size will fit, so it works best together with size guidance.
- Can I use virtual try-on images as product photos?
- Shopper try-on images are made for one shopper and are not meant for the catalog. For product pages, brands use catalog rendering tools, which put the garment on a model and output images at catalog quality.
- How does AI virtual try-on work?
- A generative model takes an image of the person and an image of the garment and renders the garment onto the person, adjusting folds, shadows and fit to the pose. Augmented reality try-on for glasses or shoes instead overlays a 3D model on a live camera image.
- Is virtual try-on the same as an AI fashion model?
- No. Virtual try-on puts a product on a given person, often the shopper. An AI fashion model is the person in a generated product image. Catalog rendering combines both: a garment rendered onto a model for the product page.