A scanned PDF is just pictures of pages: you cannot select, copy or search the words. This PDF OCR tool reads every page with the Tesseract 5 OCR engine and gives you the text, page by page, ready to copy or download as a .txt file. Tick Also make a searchable PDF to get a new PDF that looks like the scan but has a hidden text layer, so you can search it with Ctrl+F and copy text from it in any PDF reader.
Pages that already contain real text are detected automatically and their text is taken directly, so mixed PDFs are fast and exact. Each scanned page is rendered at 300 dpi, straightened and cleaned before reading. It reads 20 languages, including English, Khmer, Chinese, Thai and Vietnamese, and runs in your browser by default, so the PDF is not uploaded.
How to use the PDF OCR tool
- Click Choose PDF files or images or drop your scanned PDF onto the box. Images work too.
- Pick the Text language of the document. Under More options you can choose the pages and tick Also make a searchable PDF.
- Click Extract text and follow the progress page by page.
- Check the result in the text box; the Confidence view marks words that may be wrong.
- Copy the text, download it as a TXT file, or download the Searchable PDF.
Why use ToolOrca PDF OCR
- OCR for scanned and photographed PDFs
- Searchable PDF with a hidden text layer
- Uses the existing text of digital pages
- Choose pages (e.g. 1-3, 5)
- 20 languages including Khmer and Chinese
- Automatic straightening and clean-up at 300 dpi
- Server option for long documents
- Free, no sign-up, no watermark
Text or searchable PDF?
Choose plain text when you want to reuse the words: paste them into a document, a translation tool or an email. Choose the searchable PDF when you want to keep the original look of the pages, for example for an archive, a contract or a book: the scanned pages stay exactly as they are and the recognised text sits invisibly on top, so search, copy and screen readers work.
The searchable PDF is made on your device (the option is not available in server mode). Its pages are stored as good-quality JPEG images, so the file may be larger or smaller than the original scan.
Tips for better results
Scan at 300 dpi in greyscale if you can; colour is rarely needed for text and 600 dpi only makes files bigger. Make sure pages are not cut off and the text is not tilted much (a few degrees are straightened automatically).
For long documents, try a few pages first with PDF pages to check the language and layout settings, then run the whole file. If your phone is slow, switch to On our server, which reads pages with the same engine natively.
If a PDF made from Word gives strange characters instead of text, its fonts have no proper character map. Untick use the text already in the PDF so the pages are read by OCR instead.
Frequently asked questions
Is PDF OCR free?
Yes, free with no sign-up, no watermark and no page limit when it runs on your device. Server mode has a fair-use limit.
Is my PDF uploaded?
Not by default. With On this device (the default) the OCR engine runs in your browser and your files never leave your phone or computer; only the engine and the language data are downloaded once from a public CDN. If you choose On our server, each page is cleaned up in your browser and the prepared page image is uploaded over an encrypted connection, read by the same engine and deleted right after the text is sent back.
How long does it take?
In a modern browser a typical page takes a few seconds; the first run also downloads the engine and language data once. Pages that already contain text are instant. On a slow phone, server mode is usually faster.
Can it OCR Khmer PDFs?
Yes. Choose Khmer as the language (English is read too). The Khmer clean-up fixes the order of vowel signs and subscripts. See Khmer OCR for tips on Khmer documents.
Is there a page limit?
Not on your device, though very long PDFs take a while and use memory; you can split the work with the page range option. Pages are rendered at up to 12 megapixels.