About Extract Tables from PDF
This tool finds tables in PDFs by looking for where words line up on each page. It reads the PDF with pdf.js and shows a preview of the rows before you export. You can save results as an Excel .xlsx file (one sheet per page or one sheet for all) or as CSV. The tool turns numbers that use thousands separators into real numeric values so Excel and spreadsheets treat them as numbers. It only works on PDFs that contain real text, such as exports from Excel, bank statements, or accounting software. Scanned pages need OCR first. Everything runs in your browser, so nothing is uploaded to a server and you do not install software.
How to use it
- 1Open a PDF file in the tool (drag and drop or select file).
- 2Preview detected rows and correct any column splits or header choices.
- 3Choose export format: XLSX one-sheet, XLSX one-sheet-per-page, or CSV.
- 4Download the resulting file to your computer.
Worked example
Input: bank_export.pdf (3 pages, 50 rows per page, totals formatted like 1,234.56). Result: Option A - Excel .xlsx with three sheets named Page1, Page2, Page3, each with 50 rows. Option B - Excel .xlsx with one sheet named AllPages containing 150 rows. The value 1,234.56 becomes the number 1234.56 in Excel.
How it works
Finds tables in a PDF from where the words line up on each page, read with pdf.js, and saves them as an Excel .xlsx file (one sheet per page or one sheet for all) or as CSVShows a preview of the rows first and turns numbers with thousands separators into real numbersWorks on PDFs with real text, such as exports from Excel, banks or accounting software; scans need OCR firstWorks in the browser with nothing to installSource: PDF 32000-1 (PDF specification)pdf.js (Mozilla) - PDF reader in the browserTabula - extract tables from PDFs. Last checked 13 November 2026.
What it is good for
- Runs entirely in your browser; no upload required.
- Previews rows so you can check columns before export.
- Exports XLSX (one sheet per page or single sheet) or CSV, with numbers recognized.
- Only works on PDFs that contain real text. Scanned images need OCR first.
- Complex layouts, merged cells or inconsistent spacing may need manual cleanup after export.
Questions people ask
Will this work on scanned PDFs or images?
No. The tool reads text positions from the PDF. If a page is image-only you must run OCR first to create a text layer, using a tool such as Tesseract or Adobe Acrobat.
Do I need to install anything or upload my file?
No. The extraction runs in your browser with pdf.js. Files do not need to be uploaded to a server, and you do not install software.
Which export should I choose, CSV or XLSX?
Choose XLSX when you want typed numbers, multiple sheets, or easier Excel workflows. Choose CSV for simple imports, scripts, or systems that only accept plain text CSV files.
How does the tool detect table rows and columns?
It groups text by Y position into rows and finds column breaks by gaps in X position or by alignment on the page. The method works best on PDFs exported from spreadsheets or accounting software.
Will numbers like 1,234 be parsed as numbers?
Yes. The tool converts numbers that use thousands separators into real numeric values so spreadsheets treat them as numbers. Locale differences in decimal separators may still need checking.
Can it join tables that span multiple pages?
You can export all pages to one sheet to combine multi-page tables, but you may need to remove repeated headers or check for split rows after export.
Help keep every tool free
FreeToolr stays free because people help it grow. Share a tool you like, tell us what to build next, or support us on Ko-fi.