Documents tool

Invoice Extractor

Pull invoice fields — number, dates, totals, line items — from PDF invoices locally, with confidence and source location for every value. Missing fields are never invented.

100% local — files never leave your deviceProcessed in your browser

Deterministic label-pattern extraction over the invoice’s own text layer. A value appears only when a pattern actually matched — missing fields stay “not found” instead of being invented.

How it works

  1. Upload a digital (text-based) PDF invoice.
  2. Run extraction — fields and validated line items appear with confidence and source page.
  3. Cross-check highlighted values against the source, then export JSON or CSV.

Good to know

Frequently asked questions

Which fields can it find?

Invoice number, invoice date, due date, seller, bill-to, currency, subtotal, tax/VAT/GST and total — plus line items that pass the math validation.

It missed a field my invoice clearly has. Why?

Layouts vary enormously. The extractor only reports pattern matches to avoid inventing values. Check the source page shown per field; unmatched means “verify manually”, never “made up”.

Does it work with photos or scans?

Digital PDFs directly. For scans, convert through OCR PDF first — the extractor reads the recovered text.

Related tools