OCR PDF
Extract text from scanned PDFs using optical character recognition
Upload PDF
Drag and drop files here, or click to browse
About OCR PDF
OCR reads the pixels of a scanned page and writes a text layer underneath them, turning an image-only PDF into one you can search, select and copy from. The visible page is unchanged — the recognised words sit invisibly behind the original scan.
Recognition runs on this machine through a WebAssembly engine. Only the language training data is fetched as a static asset — your page images are never transmitted.
How to use OCR PDF
- Load a scanned PDF. Pages that already contain real text are detected and left alone.
- Pick the language of the document — accuracy depends heavily on this, and a wrong choice is the usual reason results look like noise.
- Start recognition. Progress is reported page by page, since each page is processed in sequence.
- Download the searchable PDF and test it by searching for a word you can see on the page.
On a production
Location agreements and equipment hire notes come back from set as photographs of paper. Six months later somebody needs the one that mentions a particular address, and there is no way to search forty scanned files. Adding a text layer makes the whole archive findable without retyping a single page.
Frequently asked questions
- Does OCR change how the page looks?
- No. The scanned image is kept exactly as it was and the recognised characters are placed in an invisible layer aligned to it. What you see is the original; what you search is the text layer.
- Why is the recognition wrong on my document?
- The three usual causes are a mismatched language setting, a scan below roughly 200 DPI, and heavy skew from a photograph taken at an angle. Rescanning straight and flat helps more than any setting here.
- Can it read handwriting?
- No. The engine is trained on printed type and will produce unreliable output from cursive or block capitals written by hand. Treat handwritten pages as images and index them some other way.