PDF Table Extractor
Pull tables out of a text based PDF and download them as CSV. The tool groups the text of each page into lines by position, finds runs of lines with several columns and lines the cells up by their horizontal position. It is a heuristic, so check the result, especially for tables with merged cells or wrapped text.
Frequently asked questions
How are tables detected?
Text on the same baseline forms a line, and a gap wider than about 0.8 times the font height starts a new cell. Two or more consecutive lines with at least two cells form a table, and the columns come from the horizontal extent of the cells, so left aligned and right aligned columns both work.
Does it work on scanned PDFs?
No. It reads the text layer, so a scanned page without one has nothing to extract. Run OCR first and use the result.
What happens with cells that wrap onto several lines?
Each line of text becomes its own row, so a cell that wraps produces an extra row with only that column filled. Headings that span several columns can also merge columns. Check those rows in a spreadsheet after the export.
What do I get when there are several tables?
One table downloads as a CSV file. Several tables download as a ZIP file with one CSV per table, named page-1-table-1.csv and so on. Values are quoted when they contain the delimiter, quotes or line breaks, and lines end with CRLF.
Can I limit the pages?
Yes. Enter pages like 1-3,5 to read only those pages, or leave the field empty for all pages.
What are the limits?
PDFs up to 10 MB and 200 pages, up to 200 tables, 20,000 rows in total and 100 columns per table.