Quoting a clause accurately
Pulling exact wording from an agreement avoids the transcription errors that creep in when retyping, and gives you something searchable to paste into notes or an email.
Browser-local PDF tool
Use PDF to Text
The text layer is read on this device and never uploaded. Scanned pages have no text layer, so they are reported rather than guessed at.
Only real text is returnedA scanned page holds a picture of words, not words. Pages without a text layer are named in the result instead of being silently skipped.
Explore pdf tools
Find related browser-local tools for nearby tasks without starting another search.
Browse all pdf toolsAbout PDF to Text
Most people reach for this because selecting text in a reader produced nothing usable: the selection jumped columns, the copy came out as one unbroken wall, or the reader refused to select at all. A PDF stores glyphs at coordinates rather than sentences in order, which is why naive copying scrambles anything more complicated than a single column. This tool reads those coordinates, reconstructs lines and reading order, and hands back plain text. It runs inside your browser, so the contract, statement, or medical letter you are pulling quotes from is never transmitted to anyone.
What happens to the PDF
- Choose a document. Its text layer is read one page at a time, with progress shown, so a long report does not look frozen while it works.
- Positions become lines. Each fragment carries its own coordinates, and fragments sharing a baseline are stitched back into a single line rather than being emitted as separate scraps.
- Layout is examined before ordering. If a wide blank corridor runs down most of a page, the page is treated as two columns and read down one side before the other. Sorting purely by vertical position is exactly what makes a two-column academic paper come out as alternating half-sentences.
- The result is shown, ready to copy or save as a .txt file. Everything happens in this tab, and closing it discards both the document and the extracted text.
Common pdf to text jobs
A two-column paper
An academic PDF that copies as alternating half-sentences comes out in column order instead, with a note confirming that a multi-column layout was detected.
Checking whether a scan is searchable
If you are unsure whether a file is a real document or a photograph of one, run it here. A blunt no-text-layer answer settles the question faster than hunting through a reader's search box.
Check before relying on the PDF
- A scanned document contains no text to find. Photographed or scanned pages hold an image of writing, and no amount of reading coordinates will produce characters from pixels. Rather than returning a convincing blank result, this tool says the page has no text layer. Turning pictures of words into words requires optical recognition, which this tool does not run.
- Mixed documents are handled page by page rather than all or nothing. A report with two scanned inserts still yields the text from every other page, and the result names the pages that produced nothing so you know precisely where the gaps are.
- Reading order is reconstructed, not read from the file. Documents with sidebars, pull quotes, footnotes, or floating captions can interleave, because nothing in the format records the order a human would read in. Column detection covers the common two-column case and the result warns you when it fires.
- Some documents carry fonts with no character mapping. The glyphs draw correctly on screen but no letter is recorded behind them, so extraction produces nonsense. This is a property of how the file was generated and is detected and reported rather than handed over silently.
- A password-protected PDF is decrypted in the browser once you supply its password, so extraction works normally. There is no way to read one without the password, and this tool does not attempt to guess or crack it.
Questions about PDF to Text
Why did I get nothing back from my scan?
Because there was nothing to get. A scan stores an image, not characters. The tool reports this rather than returning an empty box that looks like a failure, so you know the file needs optical recognition instead.
Is the layout preserved?
Lines and reading order are preserved; visual layout is not. You get flowing text rather than a reproduction of columns, tables, and spacing. Tables in particular flatten into lines, because plain text has no way to express a grid.
How large a document can it handle?
Up to 60 MB and 300 pages. Reading a text layer is far cheaper than rendering pages, which is why this limit is much higher than the one on tools that convert pages to images.
Can I get the text as a file?
Yes. The result can be copied to the clipboard or saved as a UTF-8 .txt file created locally in this tab, so accented characters and non-Latin scripts survive intact.
Can it read a password-protected PDF?
Yes, if you know the password. Choose the file, and if it is encrypted a password box appears; enter the password and the text comes out normally. RC4 and AES-encrypted documents are all handled. The password is used in this tab and is never stored or transmitted.
Related tools
PDF to Markdown
Turn a PDF into structured Markdown in your browser. Headings are inferred from font size, paragraphs and lists are rebuilt, and no file is ever uploaded.
Split PDF
Split a PDF into page batches, extract a custom range or save each page separately. Keep text and page dimensions with browser-local processing and no upload.
Compress PDF to 200KB
Reduce a PDF below 200KB entirely in your browser. Lossless-first compression, exact-size search, no file upload, and a local download.
PDF to JPG
Turn PDF pages into JPG or PNG images in your browser. Pick 72, 150, or 300 DPI, get one image per page, and upload nothing to a server.
Remove PDF Metadata
Read, edit or clear the title, author, subject and keywords stored inside a PDF. Runs in your browser with no upload.
TXT to PDF
Turn a plain text file into a paginated PDF in your browser. The result holds selectable text, not a screenshot.
DOCX to Text
Pull the plain text out of a Word document in your browser. No formatting, no hidden characters, nothing uploaded.