OCR PDF
Extract text from scanned PDFs with OCR in your browser — English, Urdu, Arabic and more. Per-page results, .txt download. Free, private, no signup.
About the OCR PDF
This OCR PDF tool turns scanned PDFs into editable text without uploading your document anywhere. Each page is rendered at high resolution and read by the Tesseract OCR engine running in your browser via WebAssembly — the same open-source engine used in countless document pipelines, working entirely on your device.
It supports English, Urdu, Arabic, Hindi, Spanish and French. Pick the language of your document before starting; the first run downloads that language's recognition data (a few megabytes) and caches it for next time. Progress is shown page by page, and you can process up to 20 pages per run to keep your browser responsive — split larger PDFs first and run them in batches.
Because everything is local, this is a safe choice for contracts, ID documents, bank statements and other sensitive scans. You get a combined, editable text box plus per-page sections, a copy button, and a .txt download. Accuracy depends on scan quality — clean, straight, 300-dpi scans give the best results.
How to Use the OCR PDF
- 1Upload a scanned PDF (up to 20 pages per run).
- 2Select the document's language — English, Urdu, Arabic, Hindi, Spanish or French.
- 3Click Extract Text and watch the per-page progress.
- 4Edit the combined text if needed, then copy it or download it as a .txt file.
Frequently Asked Questions
Is my PDF uploaded to a server for OCR?
No. The pages are rendered and recognized entirely in your browser using the Tesseract engine compiled to WebAssembly. Only the language data file is downloaded (once, then cached) — your document never leaves your device.
Does it support Urdu?
Yes — select Urdu from the language menu. Urdu OCR is harder than Latin scripts because of the Nastaliq style, so results are best with clean, high-resolution scans of printed (not handwritten) text.
How accurate is the text recognition?
On clean, straight scans of printed text at around 300 dpi, accuracy is typically very good. Skewed pages, low-resolution photos of documents, decorative fonts and handwriting reduce accuracy — the output box is editable so you can fix mistakes before copying.
Why is there a 20-page limit?
OCR is CPU-intensive and each page can take several seconds. Capping a run at 20 pages keeps your browser responsive. For longer documents, split the PDF with our Split PDF tool and OCR the parts separately.
My PDF already has selectable text — do I need OCR?
No. If you can select text in your PDF viewer, use the PDF to Markdown tool instead — direct text extraction is instant and 100% accurate, while OCR is meant for scans and image-only PDFs.