Copying text out of a PDF by hand rarely goes well. Selections jump between columns, line breaks land in the middle of sentences, words split by hyphens stay split, and long documents mean dozens of copy and paste rounds. This converter reads the text of the whole PDF at once and gives it back as clean plain text you can paste into a document, an email, a spreadsheet or a content tool.
It uses pdf.js, the PDF engine inside Firefox, and runs in your browser. The PDF is not uploaded anywhere, which matters for contracts, reports and anything with personal data in it.
How to convert a PDF to text
- Drop the PDF on the box at the top of this page, or click it and choose the file. The text of every page is read straight away.
- Look at the numbers above the text: pages read, pages with text, words and characters. If some pages have no text, a note names them (more on that below).
- Adjust the options. The text updates instantly, with no need to read the file again:
- Mark page breaks puts a line such as
--- Page 2 ---before each page. - Join words split by a hyphen at line end turns "inven-" plus "tory" on the next line back into "inventory".
- Join lines into paragraphs removes the line breaks inside each paragraph, so text flows when you paste it into a document.
- Mark page breaks puts a line such as
- To read only part of the file, type pages such as
3-9in Pages and click Extract text again. - Click Copy text to put it on your clipboard, or Download .txt to save a file named after your PDF.
Why people convert PDFs to text
- Reusing content. Product specs from a manufacturer sheet, policy wording from an old document or a supplier's ingredient list are easier to work with as plain text than as a PDF you have to copy from page by page.
- Searching and comparing. Plain text can be searched in any editor, compared between two versions of a contract with a diff tool, or counted with a word counter.
- Feeding other tools. AI writing tools, translation apps, text-to-speech readers and spreadsheets all accept plain text more reliably than PDF uploads.
- Accessibility. Text can be enlarged, read aloud by a screen reader or reflowed on a small screen, which a fixed PDF page cannot.
- Archiving. A .txt file is tiny, opens in any program and will still open decades from now.
How the converter rebuilds the text
A PDF does not store paragraphs. It stores small runs of text, each with an exact position on the page. Two words that look side by side may be separate pieces, and a line may be written out of order. To get readable text back, the converter:
- Groups pieces into lines by their height on the page, then sorts each line from left to right.
- Puts spaces back wherever there is a visible gap between pieces, because many PDFs leave spaces out and rely on position alone. Very wide gaps, such as between table columns, become three spaces so the columns do not run together.
- Finds paragraphs by comparing the gap between lines with the usual line spacing on that page. A gap clearly bigger than normal starts a new paragraph, shown as a blank line.
- Handles rotated pages. If a page is stored sideways, or its text runs up or down the page, it works out the main text direction first so lines still come out in reading order.
The result follows the visual layout of the page, top to bottom. Headings, list items and short lines stay on their own lines, which keeps the structure easy to spot in plain text.
Choosing the options
| You want to | Page breaks | Join hyphens | Join lines |
|---|---|---|---|
| Paste into Word or Google Docs as flowing text | Off | On | On |
| Quote from a report with page references | On | On | Off |
| Check a contract line by line | On | Off | Off |
| Feed text into an AI or SEO tool | Off | On | On |
| Keep addresses, lists and table rows intact | Your choice | On | Off |
Hyphen joining is careful: it only joins when a line ends in a letter followed by a hyphen and the next line starts with a lowercase letter. "Spring-" followed by "Summer" stays as written, because a capital letter usually means the hyphen is part of the name.
Scanned pages and OCR
Some PDFs have no text in them at all. A scanned contract, a photographed receipt saved as PDF or a page exported as an image is just a picture of text. The converter can only read real text, so these pages come back empty. Instead of leaving you guessing, it counts them and names them, for example "1 page (9) has no text layer, so it is probably a scan or a photo."
For each of those pages (up to 12), a Save page as PNG button saves a 200 DPI image, which is a good resolution for text recognition. Run that image through Image to Text (OCR) to get its text. For a longer scanned file, convert all the pages at once with PDF to PNG at 300 DPI first.
Note: some scanners add an invisible text layer made by their own OCR. Those PDFs do have text, and the converter reads it, but the quality depends on the scanner's OCR. If the text looks garbled, OCR the page images again instead.
Worked examples
A 12-page supplier price list
A supplier sends a price list as a 12-page PDF and you want the product names in a spreadsheet. Drop it in, turn page breaks off, leave join lines off so each product stays on its own line, and copy the text. Paste it into column A of a Google Sheet and split each row with a formula such as =SPLIT(A1, " ", FALSE), which breaks the line at the three spaces the converter puts between columns. For bank and card statements with dates, amounts and balances, the bank statement converter rebuilds the table for you instead.
Quoting from a research report
You need three passages from a 64-page report with page references. Keep page markers on and search the text for your keywords with Ctrl+F (Cmd+F on Mac) in the result box. Each passage sits under its --- Page N --- line, so you can cite the page correctly.
Turning a PDF guide into web content
A manufacturer's 20-page care guide is the raw material for a set of help pages. Extract the text with join lines on and hyphens joined, and you have clean paragraphs to edit. Rewrite rather than paste: copied text is the original author's work, and search engines favor pages that add something. If you are building a whole content site around a topic, AM Jarvis AI Site Builder plans and writes pillar and cluster pages on your own AI key.
Comparing two versions of a contract
A client returns a revised 9-page agreement and says "only small changes". Extract both versions with page markers off and join lines on, so a reflowed paragraph does not count as a change. Paste each into a free online diff checker or the compare feature of your editor. Every changed word stands out, including the ones nobody mentioned.
On a phone
Open this page in Safari or Chrome, tap the drop zone and choose the PDF from Files or Downloads. The text appears in the box under the stats. Tap Copy text and paste it into Notes, Google Docs or an email. Download .txt saves the file to your downloads folder.
What it cannot do
- Text inside images is not read. That includes scans, screenshots and text drawn as outlines in some design exports. Use OCR for those pages.
- Two-column layouts read across. The converter reads each visual line from left to right across the full page, so on a two-column page, lines from both columns alternate. For articles and academic papers, use Crop PDF to crop a copy of the file to the left column and another copy to the right column, then extract each one; text outside the crop area is skipped.
- Tables become spaced text, not cells. Columns are kept apart with spaces, but there is no spreadsheet structure.
- Odd font encodings can produce gibberish. A few PDFs, often older ones or those made by unusual software, store characters without a map back to real letters. The text then looks scrambled in every converter, including copy and paste in Acrobat. OCR is the only fix.
- Password-protected PDFs cannot be read. Remove the password first with Unlock PDF if you know it.
Troubleshooting
- Words run together. The PDF positioned the letters very tightly. Try the same page in another viewer; if copy and paste there also joins the words, the file itself has no spacing information.
- Headers and footers repeat on every page. Running heads, page numbers and confidentiality lines are real text, so they appear once per page. Remove them with find and replace in your editor, or crop them off first with Crop PDF; text outside a page's crop area is skipped.
- The text order jumps around. The page probably has columns, side boxes or captions. Extract smaller page ranges and tidy the order by hand.
- Accents or symbols look wrong in the .txt file. The file is saved as UTF-8, which handles every language. If Excel or an older app shows odd characters, import the file and choose UTF-8 as the encoding, or open it in Notepad, TextEdit or VS Code.
Other ways to get text from a PDF
On a Mac, Preview lets you select text with the text tool and copy it, one page or one selection at a time. In Chrome, Edge or Acrobat Reader you can press Ctrl+A on a page and copy, which works for short documents but drags in headers, footers and broken line breaks. Microsoft Word can open a PDF and convert it to an editable document, which is useful when you need the layout, but it is slower and often turns lines into separate text boxes. When you just need the words, plain text from this converter is the fastest route.
Privacy and limits
The converter is free, has no daily limit and needs no account. The PDF is read in your browser and the text never leaves your device unless you copy or save it. Very long PDFs, with thousands of pages, depend on your device's memory; extract them in page ranges if the browser slows down. To count the words in the result, paste it into the word counter.
The converter only reads; it never changes your PDF. You can run the same file as many times as you like with different options or page ranges, and the stats update each time so you can see exactly how many words came from the pages you picked.