Recognize Text via OCR
Extract text from scanned PDFs and images using optical character recognition.
Combine multiple PDFs into one
Split a PDF into separate files
Reorder pages by dragging
Pull specific pages out
Pull all images out
Remove pages from PDF
Rotate pages in any direction
Stack one PDF onto another
Reduce PDF file size
Clean up unused data
Make scanned text searchable
Spot the differences between two PDFs
Add text, signatures, shapes
Add your signature anywhere
Add text or image watermark
Number pages as 1, i or A
Add sticky-note comments
Ready-to-use forms to fill online
Add a password
Unlock a PDF with its password
Black out sensitive info
Editable .docx export
Tables as .xlsx
Each page as a JPG
Each page as a PNG
Each page as a WebP
Turn JPG images into a PDF
Turn PNG images into a PDF
Combine JPG, PNG, GIF or WebP into one PDF
Convert .docx to PDF
Convert .xlsx to PDF
Convert .pptx to PDF
Turn an HTML file into a PDF
Capture a webpage
34 PDF tools in total. Pick a category above or search for one.
OCR PDF scans and photos into plain text
A scan is a picture of a page: you can read it, but you can't select, copy or search the words. Optical character recognition looks at the shapes on the page and writes out the characters they stand for.
Every page is read as an image, even in a PDF that already has selectable text, so the result follows what is printed on the page. This image to text process leaves you with words you can paste straight into an email, a document or a search box.
Reads
- Scanned and digital PDFs
- JPG and PNG photos
- GIF and WebP images
Leaves out
- Fonts and colors
- Columns and tables
- Pictures and drawings
What happens when the text is read
When you click Recognize Text, the file is read on the server in one request. Each PDF page, or the image itself, is treated as a picture, and the language you picked decides which alphabet and accents recognition looks for.
The characters it finds come back as plain text under a heading that names the language used. Saved as a file, the text takes your document's name with _ocr.txt added and is stored as UTF-8, so accents and non-Latin scripts come through intact.
Several files are read one after another in the same language. Instead of showing the text on the page, the result lists each file and packs one .txt per document, named after it, into a ZIP called recognised-text with the number of files added. A file that can't be read is listed with the reason while the others go through.
Extract text from a scanned PDF
-
Add a scan or photo
Drop a PDF or an image onto the dropzone, or click it to browse.
-
Choose the language
Pick the language the document is written in under Language before you start.
-
Copy or save the text
Click Recognize Text, then use Copy to Clipboard or Download as Text File.
Getting accurate text recognition
-
Match the language
Read a French letter with English selected and the accented words come out wrong. Choose the document's own language every time.
-
Sharp, straight pages
Blur, shadows and tilted lines cause most misread words. If the text comes out garbled, rescan the page or take the photo again straight on.
-
Printed text, not handwriting
Recognition is built for typed and printed text. Handwriting and ornate fonts are often misread or skipped.
-
One language per run
A page that mixes two languages is read in the one you select. Run it again in the second language for the passages that came out wrong.
Can you already select the text?
Then the PDF has a text layer, and PDF to Word turns it into a .docx file you can edit in a word processor.
When reading text from a scan saves retyping
-
Quoting a scanned contract
Copy a clause word for word instead of retyping it from the page image.
-
Old letters and records
Turn typed letters, certificates and archive scans into text you can file and search on your computer.
-
Receipts and invoices
Pull amounts, dates and reference numbers out of a photographed receipt.
-
Printed handouts
Get the text of a worksheet or a notice into a document you can edit and reuse.
OCR PDF questions
Still stuck? Contact us
No. It reads the text out of your file rather than adding a hidden text layer to it, so the PDF stays as it was. Paste the text wherever you need to search it, or keep the .txt file next to the scan.
Yes. Choose Chinese (Simplified), Chinese (Traditional), Japanese, Korean, Arabic or Hindi under Language, the same way as any other language.
Check it before you rely on it. Recognition can mix up look-alike characters such as 0 and O or 1 and l, especially in faint or very small print.
Not while it is locked: add one and the tool asks you to unlock it first with Remove Password Protection. The unlocked copy is saved at 72 dpi, so small print may be read less reliably than from the original scan.
Recognition found nothing it could read as letters. That happens with blank or very faint pages, photos that contain no writing, and print too small to make out. Check the file and the Language setting, then try again with a clearer copy.
Yes. Save it as PNG or JPG and drop it in like any scan. Screenshots are sharp and perfectly straight, so they usually come out cleaner than photos of paper.
Paste it with Copy to Clipboard, or open the downloaded .txt file in your word processor. From there, save it in the document format you need.