PDF OCR
Turn scanned PDFs into searchable, selectable text. Runs Tesseract 100% in your browser — nothing is uploaded.
Pro batches up to 20 scanned PDFs through Tesseract and returns searchable PDFs in one ZIP. Ad-free, still 100% in-browser.
How PDF OCR works
Upload a scanned PDF, photo of a document, or any image-only PDF.
Each page is rendered at 2× DPI and analysed by Tesseract 5 (WebAssembly). Recognized text is collected page-by-page.
Get a searchable PDF (original image + invisible text layer that Ctrl+F can find) plus a plain .txt copy of the recognized text.
Frequently asked
Which languages does it support?+
English is loaded by default. Other Tesseract languages (Hindi, French, Spanish, German, etc.) can be added on request.
How accurate is the OCR?+
Tesseract 5 achieves 90–98% accuracy on clean printed text. Handwriting and low-resolution scans drop below that.
Is the scan uploaded?+
No. Tesseract runs in a WebAssembly worker in your browser. Nothing is uploaded.
Why is the first page slow?+
The Tesseract English model (~10 MB) downloads once from a CDN and is cached by your browser. Subsequent pages are 5–10× faster.
Can I use it on 100-page PDFs?+
Technically yes, but OCR is CPU-heavy. Expect a few minutes on desktop and much longer on mobile. Split large PDFs first for best performance.