By ShinobiTools Team · Last updated: August 2026
Drop a PDF with a table in it and get an .xlsx you can sort, filter and add up — real cells, one sheet per table. Opens in Excel, Google Sheets and LibreOffice.
A table drawn with lines — a bank statement, an invoice, a price list, a report with ruled rows — is what this page is built for. The cells are read out one by one and written into a spreadsheet in the same rows and columns, so the numbers land in their own cells and you can immediately sort, filter or sum them. Every table found gets its own sheet, named after the page it came from, so a ten-page report does not end up as one long mush.
Tables held together by spacing alone, with no ruling lines at all, are harder for any tool. Those usually still come through, but check the column edges before you trust a total. And a page where the layout is really a picture — a chart, a screenshot of a table — has no cells to read; you would need OCR first.
Up to 25 pages per file. What you get back is a normal .xlsx, not a CSV dump: text stays text, and a cell that starts with = is kept as text so Excel does not try to run it as a formula.
A scan is a photograph of a table, and a photograph contains no cells — running it through this page would give you an empty spreadsheet, so we stop it at the door and say so instead.
The fix is on this same site: the OCR tool reads the picture and writes a real text layer into the PDF. Run your scan through that first, then drop the result here.
Not sure which kind you have? Try to select a number in your PDF viewer. If you can, it is text and converts directly. If your cursor draws a box instead, it is a scan — OCR first.
This tool puts rows and columns into a spreadsheet. If your PDF is a letter, a contract or an article, there is no table to find and we tell you that rather than handing you an empty file with a row per line.
For that, PDF to Word is the right page: it rebuilds the paragraphs, headings and lists as an editable document.
Your PDF is deleted from the server the moment the spreadsheet exists, and the spreadsheet itself is deleted 45 minutes later. Nothing is stored, shared or used to train anything.
There is no account, no watermark and no daily cap. Everything runs on our own servers; the ads on the page keep them online. Your original upload is deleted the moment processing finishes, the cleaned result stays available for up to 45 minutes so you can download it and is then deleted automatically, and nothing is ever sold, shared or used to train AI.
Drop it in above. You get an .xlsx with the real cells — no account, no watermark, no daily limit.
Yes — that is the whole point. Each cell of the table becomes a cell in the spreadsheet, so you can sort, filter and add them up straight away.
Each one gets its own sheet, named after the page it came from, so you can tell them apart.
Because there isn't one it can read. Either the PDF is running text — use PDF to Word for that — or it is a scan, and then our OCR has to give it a text layer first.
Usually, yes, but they are the hard case for every tool. Check the column edges before you trust a total — we would rather say that than pretend otherwise.
25 pages per file, 25 MB. That covers almost every statement, invoice and report.
The PDF is deleted the moment the spreadsheet exists; the spreadsheet is deleted automatically 45 minutes later.