Copy a table out of a PDF and it lands in one giant column; retype it and you'll lose an afternoon plus a few typos. The PDF to Excel card does the real thing: it runs Tabula — the beloved open-source table extraction engine — over your PDF, detects where rows and columns actually are, and rebuilds each table as a proper .xlsx workbook you can sort, filter, and chart.
Three extraction modes, explained simply
Grid mode is for tables with visible ruling lines — bank statements, government forms, anything printed from Excel itself. Stream mode reads whitespace-aligned tables that have no borders at all, like price lists and report line items. Auto (the default) tries grid first and falls back to stream, which is why it's the recommended starting point.
A workbook you can actually use
Every detected table becomes its own sheet, named after the page it came from, with the header row bolded and columns auto-sized. Page selection lets you grab just the pages you need — table on page 4? Ask for page 4 and skip the rest.
If a table comes out messy
Switch the extraction mode and run it again — each attempt takes seconds, and borderless layouts are the usual culprit. For photographed or scanned invoices where the table is an image rather than text, the AI route is the better tool: the Docling Table & Invoice Extractor reads table geometry visually, borderless grids and merged cells included.
Everything runs free with no signup, and like every PDFVintage tool the file is processed and forgotten the moment your download finishes — nothing sits on a server. For recurring jobs, keep a note of which mode worked for each document type: bank statements almost always prefer Grid, while exports from older reporting tools tend to favor Stream.