OCR PDF
Turn a scanned PDF into one you can search and copy from. The words on every page are read and added as a hidden text layer under the picture of the page, and you also get the text as a .txt file. It all runs in your browser, so your PDF is never uploaded.
How to OCR a PDF
OCR a PDF without uploading it
A scanned contract, an old letter or a page photographed with a phone is stored in a PDF as a picture. You can read it, but your PDF reader cannot: Ctrl+F finds nothing, and you cannot select a sentence to copy it. OCR, short for optical character recognition, reads the letters in that picture and turns them into real text.
Most online OCR services do this on their servers, which means sending them the whole document. Here it runs in your browser. Each page is drawn on your device and read by Tesseract, a well-known open source OCR engine, running inside the page. One PDF of up to 60 MB and 50 pages can be read at a time, and the progress is shown as Reading page 3 of 12 and so on.
The first time you use it, your browser downloads the text reader and the file for the language you picked from this site: about 6 MB for English and about 5 MB for Spanish. The language file is kept in your browser, so the next run starts faster. Your PDF itself is never sent anywhere.
Make a scanned PDF searchable
Drop the PDF, pick English or Español, and press Make searchable. Every page is drawn at 200 DPI, the words are read, and the page is rebuilt as the same picture with an invisible text layer placed over the words. The result looks like your scan, but in any PDF reader you can now search for a word, select a line and copy it.
The searchable file is saved as the original name plus -ocr, so scan.pdf becomes scan-ocr.pdf. Each page keeps its size and appears the way you see it in a reader, so a page that was turned sideways stays turned. Because every page becomes a new picture, the file can come out larger or smaller than the original, and links, bookmarks and form fields are not carried over.
Copy the text or save it as a .txt file
Next to the PDF you get the recognized text. It appears in a box below the result, page by page, each page under a line such as --- Page 2 ---. Copy text puts all of it on your clipboard, and the .txt file, named like scan-ocr.txt, holds the same text for Notepad, Word or Google Docs. A page where nothing could be read says "No words found on page 2." instead of leaving a gap.
If your PDF was made by a program such as Word and already has real text, you do not need OCR: PDF to text pulls that text out directly and keeps every letter exact. To change words inside a PDF, use the PDF editor.
Tips for better OCR results
OCR is only as good as the scan. Clear, straight pages of printed text in English or Spanish give the best results. Handwriting is not read reliably, and very small print, faint copies, stamps over text and pages scanned at a slant can give wrong or missing words. Always check names, numbers and amounts against the scan before you use the text.
Pages should be the right way up before you start. A page that shows sideways or upside down is not turned for you, and its words come out as nonsense, so fix it first with rotate PDF. A PDF over 50 pages can be cut into parts with split PDF and each part read on its own.