Document type
Invoice OCR
Extract invoice headers and full line items into accounting-ready JSON.
Can Mahad OCR read a Invoice?
Available now. Extract invoice headers and full line items into accounting-ready JSON. It returns invoice_number, invoice_date, due_date; vendor and buyer details; subtotal, tax, total, currency and more as JSON. Input: PNG, JPG, PDF (multi-page up to 5 pages). Limitation: Beyond 5 pages the PDF is truncated.
Fields normally extracted
- invoice_number, invoice_date, due_date
- vendor and buyer details
- subtotal, tax, total, currency
- line_items[]
Validation available
Line items are returned alongside totals so you can reconcile them in your own system.
Example response
Synthetic example — not a real document.
{
"type": "Invoice",
"values": { "invoice_number": "INV-2026-014", "invoice_date": "2026-07-01", "total": "1250.00", "currency": "USD" },
"line_items": [ { "description": "Consulting", "quantity": 10, "unit_price": 125.00, "amount": 1250.00 } ]
}Supported formats
PNG, JPG, PDF (multi-page up to 5 pages)
Common business uses
- Accounts payable automation
- Expense processing
- Supplier reconciliation
Integration
POST https://api.mahadocr.com/v1/documents/upload
Integration guide →Invoice OCR API: fields, API example, limits →Limitations
Beyond 5 pages the PDF is truncated. Mahad OCR does not guarantee that line items sum to the printed total — always reconcile.
Related features
Try it on your own document
Upload a invoice in the live demo — the file is processed in memory and not stored.