PDF to Excel
Detect real tables in a PDF from each line of text’s actual position on the page, and export them as an .xlsx spreadsheet with real, aligned columns - no OCR, no guessing from whitespace. A faster basic mode is also available for plain text-only extraction.
Conversion mode
Frequently asked questions
How does Advanced mode detect real tables?
It reads each line of text's exact x/y position on the page (the same data the PDF itself uses to place text) and looks for x-positions that repeat consistently across several rows - that's a real column boundary, not a guess from whitespace. Rows sharing the same column pattern become one table; text outside that pattern (titles, paragraphs) is kept separate instead of being forced into a fake table row.
What's the difference between Advanced and Basic mode?
Advanced mode detects real table structure and keeps columns aligned - use it for statements, invoices, and reports with genuine tables. Basic mode is a faster fallback that puts one row per line of text, splitting columns wherever there's a wide gap - it doesn't detect table structure, but works fine for simple, mostly-text PDFs.
What happens with scanned PDFs or password-protected files?
Scanned PDFs that are just images of text won't produce any extractable text since there's no OCR step. Password-protected or corrupted PDFs will fail with an error asking you to check the file.
Is my PDF uploaded to a server?
No, this runs entirely in your browser using pdf.js to read text positions and ExcelJS to build the .xlsx file. Nothing is sent to a server.