Split a scanned PDF into its documents
You fed ten papers through the scanner and got one PDF. Upload it: Tavnit reads every page, finds where each document begins and hands them back one by one, each with its type (invoice, purchase order, receipt, packing list, bill of lading, contract…). No splitting by page count, nothing to mark by hand.
- One PDF · up to 12 pages
- 5 free documents a day
- Everything deleted after 24 h
How it works
1. Upload the scan
One PDF with several documents inside, exactly as it came off the scanner. Any order, any number of pages per document.
2. Tavnit splits and classifies it
It reads every page, detects where each document starts and assigns it a type: invoice, purchase order, receipt, packing list, bill of lading, contract, credit note or other.
3. Download each document
You see the list with page ranges and types. With a free account you take each PDF separately, or all of them in a zip.
Common questions
How does it know where one document ends and the next begins?
The AI looks at all the pages together, the way a person would: changes of header, of company, of page numbering, of layout. It does not count pages or look for markers, which is why it works when one invoice runs two pages and the receipt is a single one.
Which document types does it recognise?
Here, eight: invoice, purchase order, receipt, packing list, bill of lading, contract, credit note and other. With an account you define your own types by describing them in words, and for each one you decide where it goes: to a Flow that extracts its data, to an email or to a Collection.
What does the free version include?
5 documents a day, with the full result on screen. Downloading the separated PDFs takes a Tavnit account, which has no cost.
How many pages can I upload?
Up to 12 pages per PDF in the free version, up to 4 MB. With an account there is no page limit and you can email scans straight to the Splitter.
Does it work with poor scans or mixed languages?
Yes. It handles skewed, greyscale and stamped scans, and documents in English, Spanish or other languages mixed in the same PDF. If a page cannot be read it is marked as unclassified rather than given a made-up type.
What happens to my file?
It is processed in a Tavnit account dedicated to the free tools and used for nothing else. After 24 hours the PDF and the separated documents delete themselves, automatically; we only keep the fact that a run happened, without the file or its contents. If you want it gone sooner, email support@tavnit.io.
Why does the download need an account?
Because the free version is meant for trying it out. With an account you can split every scan, send each document to its Flow and receive the extracted data in your ERP or by email.
How do I make it automatic?
Create an account, build a Splitter with your document types and forward scans to its email address, or connect it through the API. Every document it cuts goes to its Flow on its own, and from there to your ERP, Google Sheets or an inbox.