AI virtual try-on for WooCommerce, without the hype.

What virtual try-on is, in one paragraph

A shopper uploads one photo of themselves — or of a room — and an AI model renders your product onto it. They see the dress on their own body, the sofa in their own living room, the glasses on their own face. It answers the question that stops most online purchases before they happen: will this look right on me?

It is worth being blunt about what this is not. It is not a measurement tool and it will not tell anyone their size. It renders a plausible image, not a fitting.

Which stores it is worth it for

Try-on earns its place where the buying decision is visual and the product is worn or positioned. In practice that is three groups:

• Fashion and clothing. The garment rendered on the shopper's own photo. The highest-value case, because clothing has the worst return rate in retail and the most "I could not picture it" abandonment.
• Accessories and jewellery. Bags, watches, glasses, hats. Small items where the shopper's real question is proportion — how big is it actually, on a person.
• Home decor and furniture. The item placed into a photo of the buyer's own room, either into an empty space or replacing what is already there.

If your catalogue is none of those, this is not the feature to add. A try-on button on a bag of coffee is noise.

The part most guides skip: the photo

The single biggest cause of a disappointing try-on is a photo that does not contain the part of the body the product goes on. Ask an AI for trainers on a picture cropped at the waist and it has to invent the legs — and invented legs come back distorted.

So the framing has to be stated before the shopper uploads, not after they are disappointed. Roughly:

• • Head and shoulders — glasses, earrings, necklaces, hats.
• • Head to waist — tops, jackets, dresses shot from the waist up.
• • Head to toe — footwear, full-length dresses, complete outfits.

Whatever tool you evaluate, check that it tells the shopper this up front. Mirrolo shows the requirement on the upload screen and updates it when a product is added that needs more of the frame.

What it costs, and who you pay

Every generated image costs real money in GPU time, so treat "free unlimited AI try-on" as a claim to check rather than a feature. The honest question is not whether it costs, but who you are paying and whether you can see the meter.

There are two sane models. Either you buy credits from the vendor, or you connect your own key to the AI provider and the generations bill directly to your account there. The second is worth looking for — it means the vendor cannot mark up your compute, and you can see exactly what you spend.

Mirrolo does both: credits if you want it simple, or your own fal.ai key if you would rather pay the provider directly. On the free plan the second option means you buy nothing from us at all.

Privacy is not an afterthought here

You are asking customers to upload photographs of their own bodies to your store. Before you install anything, get a straight answer to three questions:

• • Is the uploaded photo stored, and for how long?
• • Is the result stored, and who can reach it?
• • Is any of it used to train a model?

With Mirrolo the shopper's photo is relayed for the generation and is not retained afterwards; the result is kept only long enough to view and download. If a vendor cannot tell you this plainly, that is your answer.

Setting it up

1. 1. Install the plugin. In your WordPress admin: Plugins → Add New, search for Mirrolo, install and activate. Installing from the directory keeps it updated automatically.
2. 2. Connect the store. One click takes you to tryvue.io to create a free account and brings you back activated. There is no licence key to copy.
3. 3. Map your categories. Point each product category at the niche it belongs to, so the right mode loads for the right product without you tagging anything by hand.
4. 4. Place the button. Before or after Add to Cart, or anywhere you like with the [mirrolo_button] shortcode. There is also a full-screen try-on page with its own URL, which is what you link from an ad or a newsletter.
5. 5. Set a daily cap. Do this before you launch, not after. A public button that spends money needs a ceiling — Mirrolo defaults to twenty try-ons per visitor per day and a short burst limit on top.

What to expect, honestly

Results are strongest on plain garments photographed straight on, and weakest on complex patterns, sheer fabrics and unusual poses. A shopper in a badly lit selfie gets a badly lit try-on. This is a tool that makes a good product photo persuasive on a real person — not one that rescues a poor catalogue.

Set it up on your ten best-selling products first, watch what shoppers actually generate, and widen from there.

Questions merchants ask first