How to extract data from invoices automatically with an OCR API

21 Jun 2026·Mahad OCR·3 min read

Manually keying invoices into accounting software is slow and error-prone. A modern OCR API reads an invoice image or PDF and returns clean, structured data — vendor, invoice number, dates, line items, subtotal, tax and total — in seconds.

What gets extracted

• Header fields: vendor name, invoice number, invoice date, bill-to.
• Line items: each row's description, quantity and amount.
• Totals: subtotal, VAT/GST (rate and amount), and grand total.

Each field comes with a confidence score, so you can auto-accept high-confidence values and send anything uncertain to a human for a quick check.

How it works with Mahad OCR

Upload the invoice (image or PDF, single or multi-page) to one endpoint and poll for the result. The engine detects that it's an invoice and extracts the relevant fields and line items as JSON.

curl -X POST https://api.mahadocr.com/v1/documents/upload \
  -H "X-MahadDoc-API-Key: YOUR_KEY" \
  -F "file=@document.jpg"

# poll the result
curl https://api.mahadocr.com/v1/documents/DOC_xxx \
  -H "X-MahadDoc-API-Key: YOUR_KEY"

The response includes document_type: "invoice", a fields array with confidence per field, and a line_items array — ready to push straight into your ledger.

Why an API beats a desktop scanner

• Works on photos, scans and PDFs — any source.
• Reads any language, including Arabic.
• Costs cents per document, not minutes of staff time.
• Catches tampered or duplicated invoices automatically.