From photo to spreadsheet
The photo is read by Tesseract, an open text recognition engine that runs in your browser. Then the page looks for the parts of a receipt: the shop name at the top, the date, each item and its price, and the subtotal, tax, tip and total. Everything it finds is editable, and the totals are checked against the items so you can spot a misread number.
Built for expenses
The CSV has one row per item with the shop and date on every row, followed by the subtotal, tax and total, so it drops straight into Excel, Google Sheets or Numbers. The JSON uses plain field names and numbers, ready for a budgeting script or an expenses app.
Private by design
Receipts show where you shop, what you buy and often part of your card number. Most receipt scanners upload them to a server. Here the reading happens on this device, the text reader downloads once and then works offline, and your receipts never leave your hands.
Questions
Why is a number wrong?
Faded thermal paper, creases and shadows can make digits hard to read. Fix the field, or correct the text below the details and read it again. The check under the items tells you when the sums do not add up.
Can it read receipts in other languages?
Choose the receipt language first. Common words for total, tax and change in English, German, French, Spanish and Italian are recognised, and both 12.50 and 12,50 style prices work.