Extract the text from a PDF
The document’s text, without its formatting. Paragraphs are rejoined, end-of-line hyphenation is repaired, and columns are read in the order you would read them.
Free · 100% local — nothing is uploaded- 1Open your PDF.
- 2Run the extraction.
- 3Download the text file.
How to extract the text from a PDF
Open the document, run the extraction, get a text file. What separates a good extraction from a raw copy-paste comes down to three details Vélin handles: paragraphs are rejoined instead of being cut at every line end, words broken by a hyphen at the end of a line are put back together, and columns are read in the order one reads them, not left to right across the page. Tables come out with their columns separated by tabs, ready to paste into a spreadsheet. This is the action one makes before reworking a text, handing it to a proofreader, counting characters or feeding an analysis tool. The document never leaves your device. The resulting file is plain UTF-8 text, which opens in any editor.
Copying the text of a PDF, or part of it
To take a paragraph rather than the whole document, simply drag across the text in the editor: the selection takes words, not rectangles, and the copy gives back paragraphs — words broken at a line end are rejoined, columns are read in order, a pasted table keeps its columns separated by tabs, and a paragraph straddling two pages is sewn back together. By touch, a long press opens the selection. It is what is most often missing when copying from a viewer: you get back a chopped-up text, one line per line break, that has to be reformatted by hand.
A scan holds no text at all
If your PDF is an image — a scanned document, a photograph of a page — there is nothing to extract: the extraction will give back an empty file or nearly so. Run it through character recognition first, which lays a text layer under the image; the extraction will then give what the recognition read. To keep the formatting rather than the text alone, conversion to Word is the better fit.
Frequently asked questions
Are columns read in the right order?
Yes. A two-column page is read column after column, not line by line across — the usual mistake of text extractors.
What if my PDF is a scan?
A scan holds no text. OCR gives it some; extraction comes afterwards.
Are words hyphenated at line ends rejoined?
Yes, the hyphenation is repaired, and a paragraph broken by a page break is sewn back together.
Do tables come out correctly?
In plain text a table loses its grid: the cells come out separated by tabs, which pastes back into a spreadsheet. For real cells, go through conversion to Excel.
Do footnotes get mixed into the text?
They come out where they sit on the page, that is, after the page’s last paragraph. Faithful to the document, not always convenient to read.