Virtual try-on for eyewear

What the shopper does

The try-on happens on your product page, with no app to download and no account. The shopper taps the try-on button, uploads a photo of themselves, usually straight from the phone, and compares the result with the original photo in a before-and-after slider. The try-on window says it usually takes 15–20 seconds.

For glasses and sunglasses, the AI is told to place them on the face, resting naturally on the nose and ears, and to replace any glasses the person already wears. The AI is told to keep the person’s face, hair, skin, body shape, pose, hands, clothing and background unchanged, to add only this one item, to preserve its colour, material, shape, details and logos, and to give it a realistic size relative to the person.

Every result tells the shopper that it is an AI preview, not a statement about fit or size.

Which photo works best for glasses

The examples on this page were made with Aynada’s Shopify app on our own demo store. The people in the photos are AI-generated models, not real shoppers; the try-on results are Aynada’s real output, not retouched.

For glasses, the ikas try-on window asks for the face and head from the front, sharp and well lit, and advises against a dark photo, a face turned away, or one covered by hair or a hand. Both examples above use that kind of photo. On ikas, when the browser cannot see a face in the photo, it tells the shopper before they continue.

How you switch on glasses try-on

On ikas, you install Aynada from the ikas App Store. No code and no theme edits are needed: the button places itself next to “Add to cart”. You choose which products show it, by category or for single products, so you can start with your glasses and sunglasses. A store new to Aynada on ikas starts with the button switched off on every product. Aynada for ikas has the details.

To decide what a product is, Aynada reads your category names, and the product name when the categories say nothing specific. On Shopify it reads the product type and the product name the same way. Clear category names such as “Sunglasses” or “Eyewear” give the AI the more specific rule. A product with no recognisable category is treated as clothing, so check a few products before you switch the button on everywhere.

Aynada for Shopify, the app the examples on this page were made with, is on its way to the Shopify App Store. There, the button is added to the product page as an app block in Shopify’s theme editor, and you choose which products show it. Details are on the Shopify page.

Where it falls short

A try-on image is a preview, not a measurement. Aynada does not know the shopper’s measurements, so the result does not tell anyone whether a frame is the right width for their face.

In our own tests, one result with a different pair of sunglasses added lettering to a lens that is not on the product, and it happened again on a second attempt. The AI is told to preserve the product’s details, but it does not always manage it, so look at a few results with your own frames before you switch the button on across your catalog. We do not show that result here.

For every kind of product, Where virtual try-on works, and where it is weaker explains what makes results weaker and why.