Extract text from a PDF
Get the words without the formatting, ready to paste, search or feed to a script.
- 1Add a text-based PDF.
- 2Extract.
- 3Download the .txt file.
Good things start with a file.
Drop PDFs or images here, or choose from your device.
Rendering page previews…
This PDF looks password-protected, so we can't show page previews here. Type the page numbers instead, or unlock it first.
Add at least two PDFs to merge.
This tool is still in development. You can arrange files locally; nothing is uploaded until it opens.
Files are uploaded over an encrypted, single-use link, processed on our own servers and deleted within one hour.
Need a different tool?
Try “merge”, “JPG to PDF” or “sign”. We list PDF tools we are building, not every format pair.
Layout-aware
Columns and tables are kept roughly aligned with spaces, so the text is readable rather than a single run-on stream.
Plain UTF-8
No proprietary format. Opens in any editor, greps cleanly, pipes into anything.
Honest about scans
If the PDF has no text layer, the result is empty rather than garbage. Scanned documents need OCR.
Extract text from a PDF: what to expect
A PDF stores text as positioned glyphs, not as paragraphs. Extraction reconstructs reading order from positions. Simple documents come out cleanly; multi-column layouts and tables come out aligned by whitespace, which is readable but not structured.
This is the fastest way to get a document into a search index, a spreadsheet or a language model prompt without paying for a full Word conversion. It is also the right first check before converting to Word: if the text extracts cleanly, the Word conversion will be good too.
Scanned PDFs are pictures of pages. There is no text to extract, so the output is empty. OCR is the tool for that, and we will label it separately when it ships.
Questions, answered
- Why is my output empty?
- The PDF is most likely a scan with no text layer. It needs OCR, not extraction.
- Does it keep formatting?
- No bold, headings or fonts. Line breaks and column alignment are approximated with spaces.
- What about tables?
- Cells are separated by spaces in layout mode. For structured tables you will need a table extractor, which is on the research list.
- Is Markdown output available?
- Not yet. PDF to Markdown needs heading and list detection we have not validated. Plain text ships first.
- What encoding is the file?
- UTF-8 without a byte-order mark.