Virtual Try-On for Clothes: How It Works, Shopify Apps and Try-On in Chat

Every fashion seller knows the question that comes right before a sale closes or dies: "How will it look on me?" A product photo can't answer it, and a size chart only answers half of it. Virtual try-on is the technology that tries to answer the rest: the shopper adds a photo of themselves and gets an image of themselves wearing your piece before they pay.

This guide explains what virtual try-on is, how it works, what it can and can't do about returns, and the three places a clothing brand can offer it today: on Google (free, and your products may already be in it), on your Shopify store with an app, and inside Instagram DMs and WhatsApp chats. Prices and countries were checked on the vendors' and Google's own pages on 28 October 2026.

We make TailorTalk, an AI sales agent for fashion brands that offers try-on inside the chat (in beta), and we say so where it comes up. The other tools here are not ours.

What is virtual try-on?

Virtual try-on shows a shopper how a garment might look on them without a fitting room. For clothes, most tools now work from photos: the shopper uploads a picture of themselves, and a generative AI model combines it with your product image to create a new picture of them wearing it. A "virtual fitting room" is the same idea under a retail name; some earlier ones built a 3D avatar from the shopper's measurements instead.

It is different from two things it gets confused with:

• Augmented reality try-on, the live camera filter used for glasses, lipstick and shoes, which overlays a 3D model on your face or feet in real time.
• Size recommendation, which uses measurements or past orders to suggest a size. Try-on shows the look; it does not tell anyone which size to order. Our guide to fashion recommendation chatbots covers the size side.

How does virtual try-on work?

Whichever tool you use, the flow is close to this:

1. The shopper taps a "Try it on" button on a product, or asks for a preview in a chat.
2. They add a photo of themselves, usually a front-facing, full-length shot in good light.
3. The tool takes your product image of that garment, which is why clean catalogue photos matter.
4. An AI model generates an image of the shopper wearing the garment, in a few seconds to a minute.
5. The shopper looks at it, saves it or shares it, and goes on to pick a size and pay.

The result is only as good as the two photos going in. Google's own help page for merchants says the quality of its try-on "depends on the quality of the images you provide, and on the user's photo", and that a generated image "doesn't indicate fit, suggest a size, or indicate size availability". That holds for every tool on this page.

Does virtual try-on reduce returns?

Clothing and shoes are the most returned online purchases, as Statista's chart of return rates by category shows. Coresight Research estimated the average US online apparel return rate at 24.4% in 2023, and the brands and retailers it surveyed named size and fit as the top reason (53% of them), ahead of colour (16%) and damage (10%). As reported by FashionUnited in June 2026, Coresight puts the rate at 23.4% for 2025, with nearly 70% of shoppers who returned clothes citing size and fit.

So be careful with the promise. Try-on answers "will this colour, neckline and length suit me?", which is the looks part of the doubt. It does not answer "will a medium fit me?", which is the bigger part of returns. The brands that get the most from it pair it with a clear size chart, or with an agent that asks for measurements before recommending a size. Treat any app's "cut returns by X%" claim as marketing until your own return data shows it.

Google's free virtual try-on: your clothes may already be in it

Google added try-on to Search and Google Shopping in the US in July 2025 and has expanded it since (UK and India in December 2025). Its Merchant Center help, read on 28 October 2026, lists these countries: the United States, the United Kingdom, Canada, Australia, Argentina, Brazil, Chile, Colombia, Hong Kong, India, Indonesia, Japan, Malaysia, Mexico, the Philippines, Singapore, South Korea and New Zealand. It is not offered in the UAE, Saudi Arabia or continental Europe.

What a brand needs to know:

• Any merchant with a shopping feed who is eligible for free listings is opted in automatically. You don't install anything; a "Try it on" button appears on eligible product results.
• Supported: tops, bottoms, dresses and shoes. Not supported: lingerie, swimwear and accessories.
• It shows only on free (non-sponsored) results, not on Shopping ads, and shoppers must be signed in and 18 or older.
• Your images decide whether a product qualifies: at least 512 x 512 pixels, ideally 1024 or more; one garment, shown whole, on a front-facing model with arms down, on a mannequin, or laid flat without heavy folds; nothing covering the details; sleeves down, hoods down, zips closed, tops untucked.
• You can opt out through Merchant Center support if you don't want it on your products.

If you sell on Shopify, your products reach Google through the Google & YouTube channel, so this is mostly an image job. Photos that are on a model, full length and unstyled qualify more often than editorial shots. If you only have flat lays or hanger shots, an AI photoshoot tool can turn them into on-model images first; our guide to AI model photoshoots covers how to keep embroidery and borders intact.

Virtual try-on apps for Shopify

To put a try-on button on your own product pages, the Shopify App Store has several apps built for clothing. They all work the same way (the shopper uploads a photo and gets an AI image) and they all charge by the number of try-ons a month. Here are the ones we found, with prices from each listing on 28 October 2026:

Apart from Genlook, none of these had a review yet when we checked, so install on a few products first, test it with photos of your own team in different body types, and look closely at prints, embroidery and drape before you switch it on across the store. Links: Genlook, EganAI, TryDrobe, 10.1, TryOn AI, iTryOn AI and FitSee on the Shopify App Store.

How to work out the cost

Count try-ons, not visitors. As an example, if 2% of the shoppers who see a product page try it on, 10,000 product views a month means about 200 try-ons, which fits most $20 to $50 plans. A store selling occasion wear, where every shopper wants a preview, will use far more, and the per-try-on overage starts to matter.

WooCommerce and custom stores

Most try-on apps are Shopify-only. On WooCommerce or a custom site you either use a vendor with a JavaScript widget or connect a try-on model's API yourself; FASHN, which we cover in our list of AI tools for clothing brands, has no Shopify app and is used that way. Either route needs a developer for an afternoon or two.

Virtual try-on inside Instagram DMs and WhatsApp

A lot of fashion selling never touches a product page. The shopper sees a reel, sends "price?" in the DMs, and the sale happens in the chat, which is where TailorTalk works. Our AI sales agent answers on Instagram, WhatsApp and your website chat, and it can show a try-on preview in the same thread (in beta, switched on per brand on request).

1. The shopper asks how something will look on them, or the agent offers a preview when hesitation shows.
2. They choose to send one clear photo of themselves. Nothing happens unless they do.
3. The agent takes the garment image from your catalogue, so the preview shows the exact piece you sell.
4. It sends the preview back into the chat, and the conversation carries on to size, colour and payment.

Where it helps most:

• Occasion, bridal and high-ticket pieces. Lehengas, heavy sarees, gowns and formal suits are where shoppers stall longest and a lost chat costs most.
• Decisions made with family. A preview is an image the shopper forwards to a sister or a group chat, with your product in it.
• Buyers far from your store, including diaspora shoppers ordering for a wedding back home. Brands like Mysore Saree Udyog already sell to them over chat and video calls.
• Reopening a quiet thread: "Want to see how this looks on you?" offers something instead of chasing.
• After a photo search: a shopper sends a screenshot, image matching finds the closest pieces in your catalogue, and try-on puts the best one on them. Our post on visual search for fashion brands covers that first step.

Try-on runs on the same catalogue as everything else the agent does, so the setup is the catalogue: clean garment images built in the TailorTalk Catalog app, synced from Shopify, or pulled from your own system. Our guide to AI product catalogues explains how sellers build one from a camera roll.

How to choose where to offer it

• Go where the decision happens. If most sales close on your product pages, a Shopify app. If they close in DMs or on WhatsApp, try-on in the chat. If you run Google Shopping in a supported country, fix your images for Google's try-on first, because it costs nothing.
• Start with one collection, the one with the most hesitation (often occasion wear), not the whole store.
• Check your garments. Simple silhouettes preview well; heavy drape, fine embroidery and sheer fabrics are harder, and Google excludes lingerie and swimwear.
• Price by try-ons a month, including overage, not by the plan's headline price.
• Ask what happens to shoppers' photos. Google says it stores no biometric data and doesn't train on the photos; ask any app the same question before you install it.
• Measure it: compare returns and conversion for products with try-on against the same products before.

The limits, plainly

• A preview is a picture, not a fitting. It does not measure anyone or pick a size.
• Output quality follows input quality: a clear, well-lit, front-facing photo gives a far better result than a dark, filtered or cropped one.
• Results vary by garment, and AI can get body shape, faces or clothing details wrong.
• It should always be an offer the shopper can turn down, never a step they must take to get help.

FAQs

References

• Google Merchant Center Help: How Google's Try-on tool works, and About apparel virtual try-on (read 28 October 2026).
• Google: try-on launch in the US (July 2025) and in the UK and India (December 2025).
• Statista: Clothing & Shoes Are the Most Returned Online Purchases.
• Coresight Research: A 24.4% Online Apparel Return Rate Means $38 Billion Returns (April 2023).
• FashionUnited: Sizing intelligence is strategic priority as brands prepare for AI-driven commerce (June 2026).
• Shopify App Store listings for each app in the table, read 28 October 2026.

More for fashion brands: the best Instagram DM tools for fashion brands, how to connect Shopify to Instagram, and how to get your fashion brand recommended by ChatGPT. Want try-on in your DMs? Book a demo.