PDF to Text Converter
Extract the real, selectable text from a PDF into a plain .txt file. Runs entirely in your browser — files are never uploaded to a server.
Click to upload or drag and drop
Select a PDF file
How it works
This tool reads the PDF's own internal text layer — the same data your browser uses when you select and copy text from a PDF — so it is fast and accurate for normal PDFs. If the whole file contains almost no text (about 20 characters or fewer), it assumes the pages are scans and automatically runs English-only OCR on every page instead, which takes a few seconds per page. It skips OCR whenever there is any real text, so image-only pages inside an otherwise text-based PDF come out empty.
For scans in another language, or to force OCR on every page, use Scanned PDF to Text (OCR). If your text is in a photo or screenshot rather than a PDF, use Image to Text (OCR).
What the .txt file looks like
When the PDF has a text layer, lines are written in the order the PDF stores them, one per line, and the pages are separated by a blank line, with no page numbers added. When the OCR fallback is used instead, each page starts with a marker line such as --- Page 1 ---, followed by that page's text. The file is saved as your file name followed by -extracted.txt, and the first 800 characters are previewed.
When the fallback does and does not run
OCR starts only when the whole file has 20 characters of text or fewer. A scan whose three pages each carry a footer like Page 1 of 3 has 11 characters on each page, 33 in all, which is above the threshold. The tool then treats it as a text PDF and returns only the footers. A scan with one 11-character footer is below the threshold and goes to OCR. For a scan like the first, use the Scanned PDF to Text tool, which always runs OCR.
Limits
- The PDF can be up to 100 MB. A password-protected or damaged file shows "Couldn't read this PDF".
- Text at the same height is joined with no separator, so two columns that share a baseline can run together at the join.
- The OCR fallback reads English only.
Guides
- How to Convert PDF to Word (And When It Won't Come Out Perfect) — What PDF-to-Word conversion can and can't reliably recover, and how to get the cleanest result.
- Scanned vs Text PDFs and How OCR Accuracy Works — How to tell a text PDF from a scanned one, which tool fits each, and what decides how accurate OCR is.
Frequently asked questions
Does this use OCR?
Only as a fallback. It reads the PDF's own text layer directly. If the whole file contains almost no text (about 20 characters or fewer, which usually means scanned pages), it switches to OCR on every page automatically, in English only, which is much slower than reading a text layer.
When should I use Scanned PDF to Text instead?
When the scan isn't in English, or when only some pages are scanned. This tool skips OCR whenever the PDF has any real text at all, so image-only pages in a mixed PDF come out empty. Scanned PDF to Text runs OCR on every page and lets you choose from six languages.
Will formatting like tables or columns be preserved?
No — this extracts plain text in reading order, without layout, fonts, or table structure.
Why did my scanned PDF come back with only page numbers?
The scan probably has some real text, such as printed footers, and the total is above the tool's 20-character threshold, so it did not switch to OCR. Use Scanned PDF to Text, which runs OCR on every page.
Are page breaks marked in the text file?
Not when the text layer is read: pages are separated by a blank line only. With the OCR fallback each page begins with a line such as --- Page 2 ---.
How large a PDF can I use?
Up to 100 MB. Everything runs in the browser, so the limit protects your own tab from running out of memory.