OCR & Documents
Convert a scanned table, receipt or photo to Excel
Free to useNo sign-up requiredNo watermarkRuns in your browser
A photo of a printed table, a scanned report, or a receipt has no real cells in it - just a picture of text roughly lined up. OCR to Excel recognizes the words, works out which row and column each one belongs to, and builds a real spreadsheet from it.
Column detection is the part that matters here: a naive "one line, one row of text" approach breaks the moment a value is short and lands in the wrong place. This clusters every recognized word's position across the page first, so a ragged column still lines up.
How this tool works
Upload your photo or scan
JPG, PNG, WebP or a scanned PDF, read directly in your browser.
Text is recognized automatically
A progress indicator shows what is happening, page by page for a multi-page PDF.
Check the preview
The first page's detected rows and columns are shown before you download anything.
Download Excel or CSV
The spreadsheet is created in your browser and saved straight to your device.
How it works
Recognition runs on the same self-hosted Tesseract engine as the OCR tool. For a scanned PDF, each page that has no real text is rendered as an image and OCR is run on it; a page that already has real text is read directly instead.
Every recognized word's horizontal position on the page is clustered into column centres - the same approach PDF to Excel uses on a real PDF's text - and each word is snapped to its nearest column, which is what keeps a shorter value in the right place instead of drifting into the column next to it.
Rows with at least three cells and real gaps between them are flagged as table-like, which the "only export rows that look tabular" option uses to drop headings and stray text and keep just the table.
How good the result is depends on the photo
A straight, well-lit photo of a clean table converts well. A tilted photo, a crumpled receipt, or a low-resolution scan will need tidying up afterwards - the preview shows what you would get before you download anything, and CSV is there if you would rather fix it in a plain text editor.
Worked examples
A photographed price list
Upload the photo and get a spreadsheet with each item and price in its own cell, ready to sort or total.
A scanned expense report
Rows are detected from the recognized text's layout, with numeric columns stored as real numbers so Excel can sum them.
Frequently asked questions
Is my file uploaded to a server?
No. Recognition and the spreadsheet are both built entirely in your browser, on your own device.
What is the difference between this and PDF to Excel?
PDF to Excel reads a PDF's own text layer, which already knows exact positions. OCR to Excel is for when there is no text layer at all - a photo or a scan - and has to recognize the words first.
Why do the columns look wrong?
Column detection depends on the words being roughly aligned in the original photo or scan. A tilted photo or very tight spacing can confuse it - check the preview, and try a straighter, higher-resolution photo if the columns are not right.
Is there a file size or page limit?
Yes - OCR runs on your own device rather than a server, so very large files and very long PDFs are refused with a clear explanation rather than being allowed to freeze the tab.