Key takeaways
- PDF OCR reads text out of scanned PDFs and photos directly in your browser, then gives you a downloadable .txt file.
- The first time you use it, a small language model downloads (once) and is cached by your browser for next time.
- Works best on clear, well-lit scans; very blurry or handwritten text will read less accurately.
- If a regular 'PDF to Word' or 'copy text' tool gives you nothing, that's a sign the PDF is scanned images rather than real text — this is the tool for that case.
Extract text in two steps
With PDF OCR:
- Drop in a scanned PDF or a photo (JPG/PNG) of a document.
- Press "Extract text (OCR)" and wait — multi-page PDFs are read page by page, so larger scans take longer.
- Review the text box, then download it as a .txt file if it looks right.
Getting more accurate results
- Scan at a decent resolution. A blurry low-res photo gives OCR less to work with than a clear 300dpi scan.
- Straighten crooked scans where you can. Heavily skewed or rotated pages reduce accuracy.
- Expect lower accuracy on handwriting. This kind of OCR is built for printed text; handwritten notes are read far less reliably.
Frequently asked questions
Does this work on PDFs that already have selectable text?
It can, but there's no need — if you can already select and copy text in a PDF viewer, use a text-extraction tool or [PDF to Word](/pdf-to-word) instead; it's faster and more accurate than OCR for text that's already digital.
Is my scanned document uploaded anywhere?
No — the recognition model runs inside your browser. The image never leaves your device.