OCR PDF
Recognise the text in scanned PDFs and photos: get a searchable PDF and copyable text. Six languages, and your file never leaves your browser.
What it does
PDFs from a scanner or phone camera are really just pictures: you cannot search them or copy their text. This tool recognises the writing on each page and gives you two things: plain text you can edit and copy, and a PDF that looks exactly the same but is now searchable.
Recognition runs on the open-source Tesseract engine entirely in your browser. Even the language data loads from this site, so scans of contracts, IDs or medical letters are never sent to a server. Language data downloads once on first use and is read from your device after that.
How to use it
- Drop a scanned PDF or photos.
- Pick the document language and accuracy.
- Press "Recognise text", then copy the text or download it as .txt or a searchable PDF.
When is it useful?
- Finding a clause in a scanned contract with Ctrl+F
- Copying a quote from a photographed page
- Making an archive of old scans searchable
- Adding a text layer before editing a scan with PDF → Word
Tips
- Straight, well-lit, sharp scans give the best results; fix sideways pages with Rotate PDF first.
- For small or faint print, choose High accuracy.
- If a document mixes two languages, select the second one too.
Frequently asked questions
- What is OCR?
- Optical character recognition: it turns the writing in a scanned page or photo into text you can read, search and copy.
- Will my scanned PDF look different?
- No. The page image is untouched; the recognised text is added on top as an invisible layer. The file looks the same but is now searchable and its text can be selected.
- Which languages are supported?
- English, Turkish, Spanish, Portuguese, French and German. For mixed documents you can add a second language.
- Is my file sent to a server?
- No. Recognition happens entirely in your browser; everything, including language data, loads from this site and your file never leaves your device.
- How accurate is it?
- Very high on clean, straight scans of printed text. Errors increase with blurry photos, handwriting and very small print; use High accuracy for those.
- Does it read handwriting?
- It is built for printed text. It partly works on neat block capitals, but results on joined-up handwriting are not reliable.
- How long does it take?
- It depends on your device; on a recent computer it is a few seconds per page. For long documents, pick a page range to recognise only the pages you need.