Guide
Local OCR and Searchable PDF Maker
Recognize English printed text in scanned PDF pages locally and create a new image-based PDF with an invisible searchable text layer.
Does OCR upload the PDF or recognized text?
No. The OCR worker, English language data, and PDF processing assets are served by NullKit and run in the current browser.
Opens the free tool. Your selected files stay in this browser.
Clear answers
Turn an image-only scan into a searchable review copy without uploading the document or sending its text to an AI service.
Recognize English printed text in scanned PDF pages locally and create a new image-based PDF with an invisible searchable text layer.
-
01
Does OCR upload the PDF or recognized text?
No. The OCR worker, English language data, and PDF processing assets are served by NullKit and run in the current browser.
Same-origin English OCR model
-
02
Does it preserve the original PDF structure?
No. Selected pages are rendered and rebuilt as image-based pages with an invisible searchable text layer. Keep the original file.
Searchable text layer and extracted TXT
-
03
Can it recognize handwriting or every language?
Not reliably. This initial release is for English printed text. Handwriting and other languages are outside its stated support.
Page-by-page local progress and output receipt
Real-life uses
Practical ways to use Local OCR and Searchable PDF Maker
These examples show where the tool can help and what still needs a deliberate human check before the result is used or shared.
Making a scanned report searchable
Add an invisible English OCR text layer so names and phrases can be searched while the scanned page remains the visible rendition.
Creating a text review copy
Download extracted TXT alongside the searchable PDF for correction, indexing, or note-taking.
Improving access to archived scans
Create a practical working copy from image-only pages while preserving the original scan separately.
Worked example
Making an old typed meeting record searchable
The PDF contains clear English typewritten pages but no selectable text.
Before
The scan can be read visually but not searched
- The page images are the only visible content
- Names, dates, and line breaks may be recognized incorrectly
After
OCR output is tested against the scan
- Choose a suitable resolution and run local recognition
- Search distinctive phrases, copy sample passages, and compare them with every source page
What you receive: an image-based PDF with an invisible OCR text layer plus extracted text; neither is an authoritative transcription.
Four local steps
Does it preserve the original PDF structure?
- 01
Same-origin English OCR model
Choose an unsigned scanned PDF.
- 02
Searchable text layer and extracted TXT
Select the pages and OCR resolution.
- 03
Page-by-page local progress and output receipt
Run local recognition and wait for every selected page.
- 04
Same-origin English OCR model
Search, copy, and inspect the output before relying on it.
Final check
Check the result before you use it
Verification belongs in the workflow. Review the downloaded result itself rather than relying only on the preview inside the tool.
- Search for several known words on every selected page.
- Copy sample passages and compare names, numbers, punctuation, and reading order with the scan.
- Keep the original PDF because links, forms, tags, signatures, vectors, and other structure are not preserved.
Clear answers
Can it recognize handwriting or every language?
Not reliably. This initial release is for English printed text. Handwriting and other languages are outside its stated support.
- This first version recognizes English printed text and can make mistakes, especially with handwriting, columns, low-resolution scans, unusual fonts, or rotated pages.
- Pages are rebuilt as images with a hidden OCR text layer; original links, forms, tags, attachments, signatures, vectors, and accessibility structure are not preserved.
- The searchable text can contain recognition errors and must not be treated as an authoritative transcription.