Receipt scanning limitations

Line items — individual products, quantities, and prices on the receipt.

Tax, tip, subtotal, discount, and fees — breakdown fields must be entered manually.

Receipt date and time.

Payment method or card details.

Currency — the scanner does not detect the currency symbol or ISO code; you must set it manually.

Thermal (fax-style) receipts that have faded — the low contrast makes OCR unreliable.

Handwritten receipts — Vision OCR is optimised for printed text.

Receipts in languages using non-Latin scripts — recognition accuracy may be lower.

Very long receipts captured in a single photo — text near the edges or bottom may be distorted.

Crumpled, wet, or torn receipts — physical damage reduces legibility.

Backlit or glare-affected photos — reflective surfaces (glossy paper, phone screens showing e-receipts) cause OCR errors.

MileLedger uses Apple's Vision framework for on-device OCR. It does not use Apple Intelligence or any cloud AI service for receipt analysis.

If OCR cannot extract a usable value, the Merchant and Amount fields remain blank and you must enter them manually.

The receipt image is still saved as an attachment regardless of OCR success.