Scan & repair
OCR PDF
Recognize the text in a scanned PDF so you can search, copy and quote it. Recognition runs in your browser; nothing is uploaded unless you ask for a searchable PDF.
Runs on your device; uploads only to build the PDF
Drag and drop here
PDFOne PDF up to 50 MB and 100 pages
What OCR PDF does
A scanned PDF is a stack of photographs. It looks like text, but there are no letters in it, which is why searching finds nothing and you cannot copy a sentence. OCR, optical character recognition, reads the shapes in each page image and works out which characters they are.
OCR PDF does that reading on your device. You pick the document language and pages, run the recognition, and compare the result with the original page by page. Pages that already contain real text are not recognized again; their existing text is reused.
You can then save two kinds of output. Create text file writes a plain TXT of the selected pages without leaving your browser. Create searchable PDF uploads your PDF with the positions of the recognized words, so the server can place an invisible text layer over every page image; the temporary copy is deleted once the file is built.
Is this the right tool?
Good for
- Making a scanned contract, lease or court filing searchable
- Copying a passage from a scanned book chapter or journal article
- Getting plain text from a printed letter for a quote or a translation
- Preparing scanned records so desktop or cloud search can index their contents
Not the right tool for
- Paper pages or phone photos that are not a PDF yet. Scan to PDF captures, straightens and recognizes them in one go.Scan to PDF
- Scanned tables you want as spreadsheet cells. PDF to Excel recovers rows and columns, not just the words.PDF to Excel
- An editable Word version of a scan. PDF to Word runs recognition and exports the text as DOCX.PDF to Word
How to make a scanned PDF searchable
Choose a PDF
Drop one PDF up to 50 MB and 100 pages. The tool checks page sizes and existing text on your device.
Set Document language
Pick the language printed on the pages. It decides which alphabet and vocabulary the engine expects.
Choose Pages to recognize
Keep All pages, or select Custom range and enter something like 1-3, 5. You can also tick pages in the preview.
Recognize
Select Recognize text and keep the tab open. Progress is shown page by page, and finished pages are kept if you cancel.
Review
Open Text & export and step through Text from page, comparing names and numbers with the original.
Export
Select Create text file for a TXT, or Create searchable PDF for a copy of the full document with a text layer. The searchable PDF needs an upload, so it is available for files up to 30 MB.
Worked example
A 15-page scanned lease you need to search
Say your landlord sent a 15-page lease as a scan and you want to find every mention of the deposit. Open it, and the file summary reads 15 pages, Ready for OCR, meaning no page has text yet.
Leave Document language on English and Pages to recognize on All pages, then select Recognize text. The first run downloads the language data, after which the counter moves through the pages.
In Text & export, go to the page with the payment schedule and compare the amounts with the image; a faint 8 can come through as a 3. The searchable PDF uses the text as recognized, so note any misread figure. Then select Create searchable PDF. The download looks exactly like the scan, and searching for deposit now highlights each occurrence.
Tips for a better result
- Resolution matters more than anything else. Text scanned at about 300 DPI is recognized far more reliably than a fax-quality page.
- Choose the language that matches the page even when alphabets look alike. The French and German settings know accented letters that English does not.
- Turn sideways pages upright with Rotate PDF first. Text lying on its side is usually not recognized.
OCR PDF: questions and answers
How do I know if my PDF needs OCR?
Try to select a word with your cursor, or search for a word you can see on the page. If nothing highlights, the page is an image and needs OCR. This tool also tells you when you open a file: the summary reads Ready for OCR or Existing text detected.
Which languages can OCR PDF recognize?
English, Spanish, French, German, Italian, Portuguese, Russian, Japanese and Simplified Chinese. Choose one language per run, matching the main language of the pages. The language data downloads to your browser the first time you use it.
Does OCR change how my PDF looks?
No. The searchable PDF keeps every original page image as it is and adds the recognized words as an invisible layer positioned over them. Readers see the same scan; search and text selection work on the layer.
Why are some words wrong in the recognized text?
OCR infers characters from their shapes, so low resolution, tilted pages, stains, tight letter spacing and unusual fonts all reduce accuracy. Similar shapes are the usual casualties: 0 and O, 1 and l, rn and m. Check names, amounts and reference numbers against the page image.
Can OCR PDF read handwriting?
Not reliably. The engine is trained on printed and typed characters, so neat block capitals sometimes work and cursive rarely does. For handwritten notes, plan to retype the parts that matter.