Vision

Match a photo to a catalog, read a receipt, check a document.

Coming soon. The API is not live. This page is the product shape so you can plan against it.

vision is a workbench tool for image understanding as a bare API. Match a shopper’s photo to SKUs. Read a receipt. Check that a document is the document it claims to be.

Bare REST, same as Identify. Usable with zero agents, or as a patch inside one.

What it will do

• Visual match against a catalog you provide
• OCR / field read on receipts and cards
• Document checks (type, completeness)

What it will not do

It will not place the order or file the claim. It returns a match or a read. Your agent, or your code, acts.

The reference page will appear under API Reference → Vision when the endpoint ships.