How to extract a table from PDF to Excel

price_list_2026.pdf: 9 pages, item codes, prices, stock counts. Typing it by hand is miserable. Copying from a PDF viewer into Excel gives you paste soup, because PDF stores text blocks on a page, not a spreadsheet. It looks like a table. Internally, maybe not.

The form below is PDF → Excel. It sends the PDF to server-side LibreOffice and returns an .xlsx. The limit is 50 MB. If the PDF has a real text layer, the result can be useful. If it is a scan, do not expect magic.

Why PDF tables are awkward

Excel stores cells: A1, B1, formulas, data types, column widths. PDF stores a page. It knows where to draw the word Total, where to draw a line, where to place the number 12,480. The relationship "this is one table row" may not exist at all.

So PDF to Excel is usually reconstruction, not a clean file save. The converter looks at text coordinates and tries to infer structure. Sometimes it guesses well: simple table, consistent columns, normal spacing. Sometimes it does not: wrapped product names, merged cells, tiny notes, two tables sitting next to each other.

Scans are their own problem. If the PDF is made of images, there is no text to put into Excel. You need OCR first: Tesseract, Acrobat, ABBYY, whatever is acceptable for the document. This tool does not run OCR. It works with PDFs that already contain text.

What you get in real life

In good cases you get an XLSX that needs light cleanup: rename the sheet, adjust column width, check totals, remove empty rows. Still much faster than typing 300 rows from scratch.

In bad cases Excel opens, but the table looks like emergency copy-paste. Headers moved, numbers became text, some rows split in two. That can still be useful if you only need raw text. If the spreadsheet will be used for money, stock or reporting, check it before trusting it.

My rule is simple: let automation make the draft, not the final numbers. After conversion, inspect the first page, the last page and a few middle rows. If the PDF had a VAT column, it should still be its own column, not glued to the total.

When PDF is the wrong source

If you control the source, ask for a real .xlsx or CSV export. PDF is good for sending and printing. It is a bad data container, especially for price lists, reconciliation acts, bank statements and anything where one missing row costs more than five minutes of email.

If PDF is all you have, the converter saves the first rough pass. After that comes adult supervision: totals, separators, blank rows, wrapped names. PDF to Excel is not a button that does accounting. It is a way to avoid starting with an empty sheet.