Receipt Scanner
Uses optical character recognition (OCR) to read a photo of any receipt, then analyzes the extracted text to automatically infer the merchant name, date, and total amount — with every field kept editable before download, since automatic extraction is a best-effort guess and may not always be 100% accurate. Once reviewed, download the data as a ready CSV file for expense tracking. All processing happens entirely in your browser, with the receipt photo never uploaded to a server.
Your receipt photo is processed entirely on your device by an in-browser OCR engine, and is never uploaded to a server.
Drag a receipt photo here or choose one
JPG or PNG — you can also take a photo directly with your phone's camera
What this tool does
Reads a photo of a receipt and extracts the key fields — merchant, date and total — with the full raw text available and the result downloadable as CSV.
When to use it
- Collecting expense receipts to submit or archive.
- Entering expenses into a spreadsheet without typing them.
- Capturing a receipt's data before the ink fades.
- Organising receipts from a business trip.
How to use it
- Put the receipt photo on the drop area or choose it.
- Choose the OCR language.
- Run the scan and wait.
- Review the extracted fields, open the raw text to check them, then download the CSV.
A practical example
Input
A photo of a clearly printed receipt.
Output
The merchant, date and total extracted into separate fields, with the full text available for checking.
Limitations
- Field extraction is inferred from the recognised text, and receipt layouts vary enormously — a field that cannot be detected is reported as not detected rather than guessed.
- Always check the total and the date before relying on them; one misread digit changes the amount entirely.
- Faded thermal receipts and creased paper are among the hardest cases.
- One image at a time.
Privacy and processing
The image is processed entirely inside your browser and never uploaded — which matters, since receipts can carry fragments of payment details.
Technical notes
- This builds on the same OCR engine as the other extraction tools here, then applies field inference to its text.
Frequently Asked Questions
No — extraction relies on a best-effort analysis of the OCR text and can occasionally be wrong, especially with unclear receipts or unusual layouts. That's why every field stays manually editable before you download.