Skip to content
Live

Document reader (OCR, live)

Upload an invoice, receipt or form and an AI vision model reads it and extracts the fields. Nothing is stored.

Drop an invoice, receipt or form here
PNG, JPG or WebP, up to 4 MB

Processed in memory by an AI vision model. Your file is not stored. Reading can take up to 20 seconds.

Extracted fields will appear here.

How it works

  1. 01The page is scanned and text is recognised.
  2. 02Fields (supplier, date, total) are extracted.
  3. 03Rules validate them; uncertain items go to a person.

What the real version needs

  • An upload endpoint with size and type limits
  • A vision model (Llama 4 Scout on Workers AI)
  • No storage: files are processed in memory
About our OCR & document AI service

Want this for your business?

Tell us the process or product and we will show you how a version of it would work for you.

[email protected]