Virtual try-on for Shopify: live AR or try-on videos
Virtual try-on on Shopify comes two ways: a live camera app on the product page, or AI try-on videos you make ahead and add to the product's media.

Virtual try-on on a Shopify store comes in two kinds: a live app on the product page, or a video made ahead of time and added to the product's media like any other clip. What is virtual try-on? explains how the two technologies differ; this page is about where each one lives in a Shopify store.
Shopify's facts come from its Product media types and Adding product media pages; Sume's from the Format catalog and Format API docs. All were read on 2026-09-28.
What kinds of virtual try-on can a Shopify store use?
Shopify's built-in AR works on 3D models: it uses the device camera and screen to show the product in the customer's surroundings, from any angle. That suits furniture or a bag on a table; it doesn't dress a person. On-body try-on comes from third-party apps, whose behavior their own listings describe. The third option needs no app on the storefront at all:
| Storefront option | Where it lives in Shopify | What you upload |
|---|---|---|
| 3D model with AR (Shopify) | Product media, viewed in the customer's own space | A GLB or USDZ file up to 500 MB |
| Live try-on app | An app installed on the theme | Whatever the app asks for |
| Generated try-on video | Product media, next to the photos | An MP4 you generate from a model photo and a product photo |
How do I make try-on videos for my product pages?
Sume's catalog has two Formats that make this video from photos: sume-virtual-try-on (Sume Virtual Try On) and sume-virtual-fitting (Sume Virtual Fitting), whose description lists “apparel PDP videos” among its uses. Which one to call, and the request field by field, are in Virtual try-on video API.
For a store, what matters is what comes back. A completed run returns primary_output_url, a media.sume.com URL that doesn't expire, so you can download the video and add it to the product yourself. Keep a note of which run made which file, so each product page shows the right garment.
- Use photos of people who have agreed to appear in your store, and review every video against the real item before it goes live.
Where do try-on videos go on a Shopify product page?
In the product's Media section, next to its photos: upload the MP4 or pick it from existing files. A product can hold up to 250 images, 3D models, or videos, and the video appears only if your theme supports video media. The step-by-step, and fixes when a video won't show, are in How to add a video to a Shopify product page.
Can I show a different try-on video for each color?
Not as variant media. Shopify says 3D models and videos can't be used for product variant media, so a try-on video attaches to the product, not to one color or size. To cover several colors, make one video per color and add them all to the product's media; the gallery takes up to 250 items. A File metafield is the other place a video can live, covered in Shopify product metafields for video.
What can't generated try-on do on Shopify?
It isn't installed on your storefront: you add each finished file to the product yourself, and every visitor sees the same clip. Both Formats make video, not still images; the other limits of generated try-on are in What is virtual try-on?.
Sources
Related posts
More in Use cases
- What are faceless videos? The formats and how each is made
Faceless videos tell a story without the creator on camera: voiceover over B-roll, on-screen text, illustrated scenes, or an AI presenter.
- What is a hero video? The website loop and the brand film
A hero video is the short, usually silent, looping clip at the top of a web page. In campaign planning, the hero video is the flagship film instead.
- What is a packshot? Meaning, examples, and use in AI
A packshot is a clean, evenly lit photo of a product, often in its packaging, on a plain background. AI tools use one as the product's reference.
- What is creative automation? Templates, data, and review
Creative automation makes marketing assets by running one template over changing data, with people approving results instead of building each one.
Written by Sume