Extract Text from PDF

When do you need to extract text from a PDF?

When you want a file's words rather than its layout: copying a clause from a regulation into your thesis, pulling invoice numbers and amounts from a statement into a spreadsheet, or feeding course notes into a translation tool. Instead of copying page by page in a PDF reader, the tool gives you all the text in one plain-text file.

What exactly the tool extracts

A PDF may carry a real text layer, as files exported from Word, apps and websites do, or it may be nothing more than pictures of scanned pages. This tool reads only the existing text layer, using pdf.js. It does not recognise letters inside images (there is no OCR).

• The result is extracted_text.txt, encoded in UTF-8.
• Each page starts with a separator line such as --- Page 1 ---.
• Pieces of text within a page are joined with spaces, so line and paragraph breaks are not kept, and tables become a run of words and numbers.
• The preview shows the first 6,000 characters; the downloaded file contains everything.

How to extract text from a PDF

1. 1
   Click "Select files" or drag a PDF onto the box.
2. 2
   Click "Process Now".
3. 3
   Read the text in the preview below the buttons and check that it is readable, not empty.
4. 4
   Click "Download Result" to save the text file, then open it in any text editor or copy what you need.

If the result is empty or jumbled

• Page separators with nothing under them: the file is scanned, meaning images only. You need optical character recognition; our article What is OCR? explains how it works.
• Arabic letters split apart or in reverse order: some programs store Arabic inside the PDF as letter shapes rather than the letters themselves, and the text comes out broken. In our tests, files exported from Word and LibreOffice mostly gave clean Arabic, with small slips such as the lam-alef ligature coming out reversed.
• You need formatting and paragraphs: try PDF to Word, which tries to keep a structure closer to the original.

Limitations

• No OCR: scanned pages produce empty text.
• One file at a time.
• A file that needs a password to open can't be read; a message asks you to open it first with Unlock PDF. Files that only carry permission restrictions (they open without a password) usually work.
• Images, fonts and colours aren't extracted; only text.
• Reading order on multi-column pages can get mixed up.

Where is the processing done?

pdf.js reads the text inside your browser, and the contents of your file aren't transferred anywhere, which matters when the document is a contract or personal paperwork.

Extract text from PDF: frequently asked questions