Convert PDF Tables to Excel

PDF to Excel runs `pdftotext -layout`, looks for lines that contain two or more consecutive spaces, and writes those rows into an `.xlsx` workbook. Each detected table becomes a sheet named `Page N` or `Page N Table M`. Use it on born-digital PDFs where columns already line up in the extracted text.

A table needs at least two consecutive tabular lines. Prose, a single header row, or a scan with no text layer yields `no_content`: no workbook. That empty result is intentional, not a silent failure.

The tool ignores request parameters. There is no page-range field on this page. Columns are split on runs of two or more spaces; a single space inside a cell stays in the cell.

This is not OCR. Image-only scans have nothing for `pdftotext` to read. OCR on this site returns Markdown, which you cannot pipe back into PDF to Excel. If the words exist only as pixels, you will not get a spreadsheet here.

PDF to CSV uses the same layout heuristic. One table becomes one CSV; several tables become a ZIP of CSV files. Excel keeps those tables as sheets in a single workbook. Neither reconstructs merged cells, colored headers, or hidden sheets from the PDF.

Open the download in Excel or LibreOffice Calc and check column splits on a few rows. Tight tables without a two-space gap often land in one column. Password-protected files are unlocked for you on this page once you enter the password; API callers run Unlock first.

The two-space rule is the whole detection mechanism, so the quality of the workbook depends entirely on how the PDF was typeset. Tables produced by report generators that pad each cell to a fixed width extract cleanly. Tables set with a single space between columns, or with proportional fonts that place columns nearly adjacent, do not register as tables at all and may produce no workbook. Before you blame the tool, select a row of the PDF and confirm the text is selectable at all.

Sheet naming follows position, not content. Each detected block becomes `Page N` or `Page N Table M`, in the order pages are scanned. A report with one table on each page produces one sheet per page; a page with three stacked tables produces three numbered sheets. There is no option to name sheets after a header cell, and there is no combined single-sheet output.

Values are typed as text, exactly as they appeared in the PDF's layout-preserved extraction. Numbers are not converted to numeric cells, so leading zeros survive and currency symbols and thousands separators stay in the string. That is usually what you want for account numbers and postal codes, and mildly annoying for a column of amounts you intend to sum. Either convert the types in the spreadsheet, or copy the column and use the destination's text-to-columns or value conversion.

Rows are not de-duplicated and repeated headers are not removed. When a table breaks across a page boundary, the header row often repeats on the next page and the continuation becomes its own sheet. Merged cells are not reconstructed; a cell that spanned two columns in the PDF simply has its text in the leftmost column with the right column empty. Multi-line cells become separate rows rather than a wrapped value.

Uploads are temporary. We do not keep a workbook library of your documents.