Turn scanned and image-only PDFs back into searchable, copyable text — read entirely in your browser, nothing uploaded.
Nothing leaves your browser — your files are processed on your own device and are never uploaded to any server. OCR runs on your device, so the first run takes a few seconds while the reading engine loads.
OCR starts as soon as you choose a file. English printed text only — handwriting and very blurry scans will not read well.
pages read
How to use this tool
Drop the file in or click to pick it. It never leaves your device — OCR runs in your browser.
The first run loads the English reading engine (a few seconds); then every page is read in turn, with progress shown.
Read through the per-page text. OCR misreads similar characters (0/O, 1/l) in poor scans, so proofread before relying on it.
Download a plain .txt file, or copy the text straight into your document.
Formula & worked example
A 2-page scanned letter reads as '— Page 1 —' followed by its text, then '— Page 2 —' and its text — ready to paste into a document or search.
Limitations — read this before you rely on it
English printed text only. Handwriting, tiny or blurry text, skewed scans, tables with merged cells and decorative fonts recognise poorly. Text-based PDFs do not need OCR at all — use a PDF-to-text tool for those, which is exact rather than guessed. Always proofread: OCR output is a best guess, not a transcription.
Frequently asked questions
Optical character recognition — reading text out of pictures. A scanned PDF is just images of pages, so OCR is what turns those pictures back into selectable, searchable text.
The reading engine and its English data (several MB) load into your browser the first time. Later runs on the same visit are faster, and nothing is ever uploaded.
No — this engine is trained for printed text. Handwriting needs a different kind of model and will mostly come out as gibberish.
No. If you can already select and copy the text, it is exact. OCR is only for scanned or image-only PDFs, and its output is always an approximation.
Yes. The PDF is rendered and read on your own device; no page is sent to any server.