Everything you'd want to know before you install.
A detailed look at how AI Visual Search by Image & Shop the Look works, why we built it the way we did, and the thinking behind the features above.
Search that starts from the image, not the words
A shopper sees a product they love — on the street, on social media, at a friend's place — but cannot name it. Text search leaves them empty-handed. With visual search they simply drop in a photo: the module compares that image against your whole catalog and surfaces the closest items, ranked by similarity score. Fewer dead-end searches, more add-to-cart.
A provider-agnostic engine: you choose the AI
The module does not lock you into one vendor. Voyage and Cohere produce a multimodal image vector directly; OpenAI and Mistral describe the image with a vision model and then embed it, which additionally unlocks Shop the Look's multi-item detection. You select the engine, enter your own key, and pay the AI as you go — no forced subscription.
Shop the Look: the whole scene becomes shoppable
From a single lifestyle photo, Shop the Look identifies every visible item — the jacket, the t-shirt, the shoes — and surfaces the matching products from your catalog, presented in distinct groups. Ideal for fashion, home decor and any world where customers buy an outfit rather than a single item.
Similar products at no extra cost
On every product page, a "visually similar products" block appears automatically. Instead of running a fresh analysis, it reuses the vector already computed when the product was indexed: the visual recommendation triggers no AI call and adds no recurring cost.
Indexing under control
Index your catalog in batches with a progress bar, right from the back office. Re-indexing fires automatically when a product is saved, and only changed images are re-encoded. The minimum similarity threshold and the result count are configurable, and cosine-similarity search stays smooth up to about 10,000 products, with no heavy external service to host.
There are no reviews yet.