Education

Virtual Try-On Explained

Sydney Stones · 2026-03-12 · 7 min read

Virtual try-on uses AI image generation to show a shopper how a garment would look on their own body. The shopper uploads a photo. The AI replaces the clothing in the image with the merchant's product, preserving the shopper's face, pose, and body shape.

The technology behind it is a diffusion model — the same family that powers Midjourney and DALL-E, but tuned for garment accuracy rather than artistic generation. Rendered Fits uses Google's Gemini Flash Image API, hosted in London.

Cost to the merchant is a monthly subscription plus a small per-try-on variable cost. At 1K resolution, Rendered Fits' variable cost is roughly £0.055 per generation. A merchant on the Starter tier (£249/month, 1,000 try-ons) pays about £0.25 per try-on all-in.

It works best for full garments where fit and drape matter: tailoring, knitwear, occasionwear, outerwear. It is less suited to accessories, jewellery, and eyewear — AR overlay is better for those.

Honest limits: complex draping, highly reflective fabrics, and extreme poses can still challenge the model. The technology improves monthly. In 2026, the output is already photorealistic for most fashion categories.

Book a demo

← Back to Journal

Frequently asked questions

How does virtual try-on work?

AI image generation replaces the clothing in a shopper's photo with the merchant's product, preserving face, pose, and body shape.

How much does virtual try-on cost?

£249–£5,000/month depending on tier, plus ~£0.055 per generation in variable costs.

What are the limits of virtual try-on?

Complex draping, reflective fabrics, and extreme poses can still challenge the model. Most standard fashion categories work well in 2026.