Split and rename scanned PDFs with OCR
A scanner produces pictures of pages, with no text to search. Here’s how to split a scanned batch and name each file from its text, with text recognition running on your own device.
Scanning a stack of paperwork usually produces one long PDF. Each page is an image, so tools that split by text find nothing to work with.
Smart PDF Split & Rename detects pages without usable text and reads them with OCR (optical character recognition) in your browser. The recognised text is then used like a normal text layer: to find where each document starts and to name the files.
Step by step
Add the scanned PDF
The tool marks the pages that need OCR: pages that are only images, and pages whose existing text looks unreliable.
Choose the language
Choose the language the documents are written in: English, Hindi, Spanish, Portuguese or French. For documents that mix languages, choose both, for example Hindi and English.
Choose what to read
Reading only the names is fastest: the tool reads the area where the name is printed on the first page of each document. Read whole pages when the split itself depends on the scanned text.
Check uncertain words
Words the OCR engine is less sure of are highlighted, with a picture of that part of the page next to the text. Correct them where needed.
Name and download
Set the file names, review them, and download the files individually or as a ZIP. The audit report shows which values came from OCR.
Settings to choose
| Setting | What to choose |
|---|---|
| OCR language | The language or languages the documents are written in, not the language of this website. |
| What to read | Names only, when you split by page count, by ranges or by hand. Whole pages, when the split depends on text in the scans. |
| Name area | Mark the area of the page where the name or number is printed. The same area is read on every document, so this works best for forms with a fixed layout. |
| Split rule | Any rule works once the text is read: a keyword, “Page 1 of N”, a changing value, or every N pages for fixed-length forms. |
Example
A four-page scan of three delivery notes. The number of each note is read with OCR from the top of its first page.
| Original pages | Split file |
|---|---|
| 1 | Delivery note 5521.pdf |
| 2–3 | Delivery note 5522.pdf |
| 4 | Delivery note 5523.pdf |
Example with made-up names.
Check before you use the files
- The wrong language gives confident but wrong text. Accented or non-Latin names are misread if the language isn’t selected. Choose every language the documents use.
- Scan quality matters. Straight pages with good contrast, scanned at around 300 dpi, are read best. Phone photos and low-resolution scans give more errors.
- Handwriting isn’t read reliably. Type the names of handwritten documents yourself when you review them.
- OCR is never treated as certain. Every name read with OCR is marked so you can check it before downloading.
- The pages themselves aren’t changed. By default the split files contain the original scanned images; the recognised text is used only to split and name them.
Your files stay on your device
Your PDFs are read, split and named in your browser. They are never uploaded, and your browser blocks the tool page from connecting to the internet or loading anything from other websites. How this works
Questions
Are my scans sent to an OCR service?
No. OCR runs in a background worker in your browser, on your device. The recognition engine and language models are downloaded from puffinkit.com when you use OCR; your scans never leave your device.
My PDF already has text from the scanner. Will it be used?
Yes, if it looks reliable. If the text looks wrong, for example full of odd characters, the tool suggests reading those pages again with OCR.
How long does OCR take?
It depends on your device and the number of pages. Reading only the names is much faster than reading whole pages. The tool shows its progress and you can cancel at any time.
More guides
- Split a payslips PDF into one file per employeePayroll software often exports every payslip in one PDF. Here’s how to split it into a separate, correctly named file for each employee, without uploading salary data anywhere.
- Split a PDF of invoices by invoice numberAccounting and billing systems often print a whole run of invoices into one PDF. Here’s how to turn it into one correctly named file per invoice, with multi-page invoices kept together.