Scanned PDF OCR (Extract Text)
Recognizes the characters in scanned, image-only PDFs and turns them into copyable, searchable text. If your PDF already has selectable text, PDF to Word is faster and more accurate.
0 B uploaded — this page has not sent your files anywhere?
Other conversion sites upload your file to a server, process it there, and say they delete it after a while. Until that happens, your file sits on someone else’s server. Here there is no upload step at all. The more sensitive the file — contracts, ID scans, medical records — the more that difference matters. See how to verify it →
Get copyable text out of a scan that will not let you select anything
A PDF made by a scanner or a phone camera stores each page as a picture, so dragging over the words selects nothing and search finds nothing. This tool renders each page as an image, runs optical character recognition on it, and hands you the text to copy, correct or save. The PDF never leaves your browser.
A closer look
Recognition is done by tesseract.js, which is loaded only when you press Start OCR. On the first run the engine and the training data for the language you picked are downloaded once, which can take about a minute; a second run in the same language reuses the prepared worker and starts immediately. Only the language data is fetched. Your PDF is not sent anywhere. Each page is drawn at double scale on a white background before recognition, so a crisp original gives noticeably better results than a blurry one.
The output is plain text with a “--- Page 1 ---” separator between pages. It appears in an editable box where you can fix mistakes, then press Copy or save it as a “_OCR.txt” file. Note that this extracts text; it does not produce a new searchable PDF. If you add several PDFs at once they are processed in turn with a separator line per file, one failure does not stop the rest, and a Stop button lets you halt part way. If your PDF already contains real text, PDF to Word is faster and more accurate.
How to use it
- Add a scanned PDF, or several at once.
- Choose the Document language: English, Korean + English, Korean, Japanese or Chinese (Simplified). The closer the match to the real content, the higher the accuracy.
- Press Start OCR. On a first run you will see the engine preparing, then progress such as “Recognizing page 3/10… 45%”.
- Read the text in the box below, correct anything the engine misread, then press Copy or Save TXT.
When it helps
Common snags
Tip Pick the narrowest language option that fits: “English” for an English-only page rather than “Korean + English”. Wider language sets give the engine more look-alike characters to confuse.
FAQ
How accurate is it?
Sharp, 300-DPI-quality scans read very well; blurry or tilted scans, handwriting and complex tables can produce mistakes. The result box is editable, so give the output a quick once-over.
Is my file uploaded?
No. Only the text-recognition feature and its language data are downloaded — the processing and recognition of your PDF happen entirely inside your browser.
Should I OCR a normal (text) PDF too?
No. If you can select the text with your cursor, PDF to Word extracts it far faster and without recognition errors. OCR is for scans where text cannot be selected.
What about documents mixing languages?
Choose "Korean + English" for documents that mix the two; for Japanese or Chinese documents, pick that language.
Recognition takes too long.
The first run is slow because language data (several MB) must download; afterwards it is a few seconds per page. For long documents, split the PDF and OCR only the pages you need.