PDF to Text Converter

Add your PDFs to the box above, choose how the text should look, and press Extract text. A moment later the words of every page appear in a text box, ready to copy, or to save as a .txt file that any program can open. Pages that are only a picture, such as scans and photos of paper, are read with OCR.

Everything happens in your browser. Your PDFs are not uploaded: the text is read on your own device, so even long reports and private letters stay with you. You can add up to 30 PDFs at once, each up to 200 MB. The only exception is a password-protected PDF: to open it, the file and your password are sent to our server once and deleted straight afterwards.

Before you start, it helps to know that there are two very different kinds of PDF, and they need different handling. Most PDFs made by a computer, such as documents saved from Word, invoices from an online shop, e-books and reports, contain a text layer. The letters are stored as real characters with a position on the page, which is why you can select and search them in a PDF viewer. Reading that layer is fast and exact: every letter comes out as it was typed.

A scanned PDF is different. When paper goes through a scanner or is photographed with a phone, each page is stored as a picture. To a computer, that picture is a grid of dots with no letters in it. You cannot select a word, search finds nothing, and a normal converter returns an empty file. To get words out of such a page, the picture has to be looked at and the shapes of the letters recognised. That is called OCR, optical character recognition.

Many real files are a mix. A contract may have typed pages and a signed, scanned last page; a report may contain screenshots with text in them. The tool checks every page on its own. With OCR on Automatic, a page that has a proper text layer is read directly, and only a page with no real text is sent to OCR. The result line tells you how many pages were read by OCR, so you know which parts deserve a second look.

A quick check before any PDF to text job: open the PDF in any viewer and try to select a word with the mouse. If a word highlights, the page has a text layer. If the whole page highlights as one block, or nothing happens, it is a picture and needs OCR.

This PDF to text converter free of charge has no sign-up, no watermark, no daily limit and no trial that runs out. You do not need an account, and you can convert one batch after another.

Because the work is done by your own browser rather than a paid server, there is no cost per page that has to be passed on to you. The limits are practical ones: up to 30 PDFs per batch, 200 MB per PDF, and up to 100 pages per PDF read by OCR, since OCR takes a few seconds per page on a normal computer.

Nothing is added to your text. There is no advert, no line about where it was made and no hidden marker. The .txt file contains the words of your PDF and, if you want them, the page markers you chose.

PDF to TXT means turning a fixed page into the simplest file there is: plain text. A .txt file has no fonts, no colours and no layout, only characters and line breaks. That makes it tiny, readable on every device made in the last forty years, and easy to feed into other programs, from a search index to a translation tool or an AI assistant.

The file you download is UTF-8 with a byte order mark, which is the form that Notepad, Word, Excel, Mac TextEdit and code editors all open correctly. Accented letters, Cyrillic, Arabic, Hindi, Chinese, Japanese and Korean characters come out as they should, not as strange symbols.

Each PDF becomes one text file with the same name, so annual-report.pdf becomes annual-report.txt. When you convert several PDFs at once, the files are packed into one ZIP for download, and two PDFs with the same name get a number added so nothing is overwritten.

As a PDF to TXT converter for many files, the tool works through the list one PDF at a time, showing the progress of each. You can add more files while you set the options, remove one with the cross next to its name, or stop the run with the Stop button and continue later.

When the batch is done, a row of buttons above the text box lets you switch between the results, so you can check each file before you download. Download all puts every .txt file in one ZIP named pdf-text.zip. If one PDF fails, for example because it is damaged, the others are not affected, and the failed one shows a short message in the list.

If you change an option after converting, such as switching from Plain paragraphs to Keep layout, the finished files are marked as waiting again. Press Extract text once more and they are read with the new settings, without adding them again.

Text in a PDF is not stored as sentences and paragraphs. It is stored as small pieces, each placed at an exact spot on the page. A line of a paragraph may be one piece or twenty. To make readable text, the tool sorts the pieces from top to bottom and left to right, groups them into lines, and adds spaces where there is a gap. Then it builds the result in one of two ways.

Plain paragraphs is the default. The lines of each paragraph are joined into one long line, extra spaces are removed, and a blank line separates paragraphs. A word that was split across two lines with a hyphen, like infor- and mation, is put back together. This is the best choice for reading, quoting, translation, AI tools, search and anything where the text should flow freely.

Keep layout does the opposite. Every line stays a line, and spaces are added so that each piece of text starts at roughly the same column as on the page. Columns of numbers stay under each other, a two-column form keeps its two columns side by side, and an invoice keeps its item, quantity and price lined up. View the result in a monospaced font, such as Notepad's default or Consolas, to see it properly.

PDF to text with OCR is what makes scanned documents useful. Under Scanned pages (OCR) there are three choices. Automatic, the default, checks every page and runs OCR only on pages that have almost no real text, which is what you want for most files. OCR every page ignores the text layer and reads each page as a picture. Never skips OCR completely, which is fast and right for files you know are typed.

For OCR, each page is drawn as an image at about 300 dots per inch, the resolution scanners use for text, and passed to an OCR engine that runs in your browser. It finds lines of letters, recognises the words, and returns them in reading order. The engine and its language data are loaded the first time you need OCR, which can take a little while on a slow connection; after that your browser keeps them, so the next time is quicker.

OCR is slower than reading a text layer: expect a few seconds per page, more on an older phone. Up to 100 pages per PDF are read with OCR. For a longer scanned book, use Split PDF to cut it into parts, or set a page range under More options and convert in steps.

Be honest with yourself about the result. OCR is very good on clean, straight, printed pages and much weaker on handwriting, faint copies, tiny print, stamps and photos taken at an angle. Even on good pages it can confuse similar shapes, such as 0 and O, 1 and l, or rn and m. Always check names, amounts, dates and account numbers against the original before you rely on them.

OCR needs to know which language it is reading, because it uses word lists and letter shapes for that language. Eighteen languages are available: English, Spanish, French, German, Portuguese, Italian, Dutch, Russian, Arabic, Hindi, Chinese (Simplified), Japanese, Korean, Turkish, Polish, Ukrainian, Vietnamese and Indonesian.

You can pick up to three at once, for example English and French for a bilingual contract, or Hindi and English for a form that mixes both. Choose only the languages that are really in the document: each extra language makes OCR slower and adds a small chance that a word is read in the wrong one.

Languages matter only for OCR. Text that is read from a real text layer comes out in whatever language and script it was written in, including languages that are not on the OCR list, with no setting needed.

Sometimes the words you want are inside a picture in an otherwise normal PDF: a screenshot in a manual, a chart with labels, a scanned receipt pasted into a report. To extract text from PDF image content like that, choose OCR every page. On Automatic, a page that already has typed text is read only from its text layer, so text inside its pictures is skipped.

With OCR every page, the whole page, typed text and pictures together, is read as one image. The typed text then also goes through OCR, which is a little less exact than reading it directly, so use this mode when the pictures really matter. If only a few pages contain such images, set a page range for those pages and convert them separately.

For single photos and screenshots that are not in a PDF at all, use Image to Text, which reads JPG, PNG and other image files the same way.

You do not always need the whole document. Under More options, the Pages box lets you extract text from PDF pages you choose. Type single pages and ranges separated by commas, such as 1-3, 7, 10-end. Leave it empty to read every page. The page markers in the result keep the real page numbers, so page 7 is still called Page 7.

A range is also the easiest way to handle a very long scanned file, a single chapter of a book, or the one table you need from a long report. If you need the selected pages as their own PDF too, Split PDF can cut them out.

A PDF to text file is easy to open, but a few details decide how well it works in other programs, and you can set them all. Between pages you can put a line such as ----- Page 4 -----, which makes it easy to find where a quote came from; just a blank line, for one continuous text; or a page break character, which some text tools and printers treat as a new page.

Line endings can be Windows (CRLF), the default, or Mac / Linux (LF). Modern editors handle both, but some older Windows programs show LF files as one long line, and some scripts and data tools expect LF. If you are not sure, keep the default.

The file is saved as UTF-8 with a byte order mark, so programs know which encoding it uses. Hyphen joining can be switched off under More options if your text contains many real hyphenated words at line ends, such as long compound names, and you want every hyphen kept.

This PDF to text online tool runs in any modern browser on Windows, Mac, Linux, Chromebook, iPhone, iPad and Android. There is nothing to install and no app to download. On a phone, tap the box to pick a PDF from the Files app, your downloads or a cloud drive, then copy the text or save the .txt file to the same place.

Because the work happens on your device, speed depends on it. A text PDF of a few hundred pages is usually read in seconds on any computer. OCR is heavier: a recent laptop handles a page in a few seconds, an older phone may take longer, and very long scanned files are more comfortable on a computer. Keep the tab open while it works.

PDF to text Windows users can skip installing a program. Open this page in Edge, Chrome or Firefox, then open File Explorer and drag your PDFs onto the box, or click the box and pick them. Press Extract text, then Download TXT. The file lands in your Downloads folder and opens in Notepad with a double click.

The defaults are already set for Windows: Windows (CRLF) line endings and UTF-8 with a byte order mark, so Notepad, WordPad, Word and Excel show every character correctly. If you chose Keep layout, view the file in Notepad or another editor with a monospaced font, or the columns will not line up.

To copy instead of saving, press Copy text and paste with Ctrl+V into any program. You can also paste PDF files onto the page with Ctrl+V after copying them in File Explorer, and start the conversion with Ctrl+Enter. The same steps work on Mac with Cmd instead of Ctrl.

Many AI assistants accept PDFs, but they work better with clean text. Some read only the text layer and miss scanned pages; some lose the order of columns; others hit a size limit sooner with a PDF than with the same words as text. Extracting the text first gives you control: you can see exactly what the assistant will read, remove pages you do not need, and paste or attach a small .txt file.

For AI tools and translation, choose Plain paragraphs. Sentences then arrive whole instead of cut at every line end, which noticeably improves translations and summaries. Keep the ----- Page N ----- markers if you want the assistant to tell you on which page it found something.

Plain text is also ideal for search. Desktop search, note apps and document systems index .txt files instantly. For a folder of scanned letters, keeping a text copy of each one turns it into an archive you can search by name, date or subject.

A PDF that is only a scan cannot be read by a screen reader at all: there is no text for it to speak. Running OCR and saving the result as text makes the content available to screen readers, braille displays and text-to-speech apps. Plain text can also be shown in any font size, colour and line spacing, which helps people with low vision or dyslexia.

Remember that OCR can make mistakes, so for important documents ask the sender for an accessible version as well. For your own documents, the best accessibility comes from creating them as text from the start rather than scanning a printout.

OCR is not perfect. On clean printed pages it gets almost every word right, but handwriting, faint or crooked scans, tiny text, unusual fonts, stamps and coloured backgrounds lower the accuracy. Always check names, numbers and dates that matter. OCR is also slower than reading text, a few seconds per page, and the first use downloads the language data.

Tables come out as text, not as a spreadsheet. Keep layout lines up the columns with spaces, which is good for reading and checking, but the cells are not separated. When you need rows and columns you can sort and add up, use PDF to CSV instead. When you need the document with its headings, bold text and images to edit, use PDF to Word.

Pictures, colours and fonts are not kept: a .txt file holds only characters. Text inside pictures on an otherwise typed page is read only with OCR every page. Up to 100 pages per PDF are read with OCR, files are limited to 200 MB, and the preview box shows the first 500,000 characters of a very long text; the download always contains all of it.

To extract text from PDF free of any upload, the tool uses your browser as the converter. The PDF is opened by a PDF reader built into the page, the text layer is read on your device, and OCR runs on your device too. Your file is not sent to our server, and the text is never stored anywhere but your own screen and downloads. You can check it yourself: after the page has loaded, the tool keeps working with your network switched off, except for the first download of OCR language data.

The one exception is a locked PDF. When a file needs a password to open, a box appears next to it. When you type the password and press Extract text, the file and the password are sent over an encrypted connection to our server once, the lock is removed, the opened file is sent back to your browser and both are deleted right away. If you would rather not do that, open the file in a program you trust and save it without a password first, or use Unlock PDF and then convert the result.

Your settings, such as the layout and OCR languages, are remembered in your own browser for next time. Your files and text are not.

When you press the button, each PDF is opened by a PDF reader that runs inside the page. For every page in your range, it asks for the list of text pieces and their positions. The pieces are sorted into lines from top to bottom and left to right, gaps become spaces, and larger vertical gaps become paragraph breaks. Plain paragraphs then joins the lines of each paragraph and repairs hyphenated words; Keep layout places each piece at its column with spaces.

If a page has almost no text and OCR is on, the page is drawn as an image at about 300 dpi and read by an OCR engine in your browser using the languages you picked. The pages are joined with the separator you chose, line endings are set, and the result is shown with its word count, character count and number of OCR pages, ready to copy or download.

Try selecting a word in the PDF first: if it highlights, OCR is not needed and Never is fastest.

Use Plain paragraphs for reading, AI tools and translation; Keep layout for statements, invoices and forms.

Pick only the OCR languages that are really in the document.

Keep the page markers when you will need to cite or find the source page.

Check every number and name in pages read by OCR; the result line tells you how many there were.

For long scans, convert a page range first to see whether the quality is good enough.

For a quick count, paste the text into the Word Counter for more detail, such as sentences and reading time.

You can select all the text in a PDF viewer and copy it. That works for a short typed page, but the lines break at every line end, columns get mixed, and scanned pages give nothing. Some office programs can open a PDF and save it as text, though they rebuild the layout first and the result varies.

Command-line PDF to text programs can do the same job well, but they need to be installed, and OCR needs a separate program with its own language files. This page does both in one place, with clear settings and nothing to set up.

Once the PDF to text step is done, you can paste the text into the Word Counter to see words, sentences and reading time, or save edited text back as a document with TXT to PDF. If you need the text with headings for notes, wikis or AI tools, PDF to Markdown keeps the structure.

For an editable document with the original look, use PDF to Word. For tables you want as rows and columns, use PDF to CSV. For photos and screenshots, Image to Text reads them with OCR. You will find all of these among the PDF Tools.