Split PDFs intelligently on your computer

OléPDF proposes explainable cuts from document identifiers, numbering resets, blank separators, document families and chapter headings. Review every boundary before creating named vector PDFs. Scanned pages can use local OCR, and the file never leaves your device. Free includes one PDF per day; the 7-Day Pass and Pro are unlimited.

100% private: everything runs on your computer, your files never leave your device.

Tools

  1. Upload a PDF and choose the detection sensitivity
  2. Review why each page may start a new document
  3. Correct any boundary and download the named parts

What Smart Split PDF does

Smart Split PDF finds where one document ends and the next one begins inside a single PDF, and proposes the cuts for you to review. It is made for bundles that mix several documents: a scan of a whole folder of paperwork, invoices and contracts exported together, or a report followed by its annexes. Every proposed cut comes with the reasons behind it, and you decide which ones to apply before any file is created.

The analysis reads the text of each page on your computer and applies fixed rules to it. It does not send the document to an AI service and no hidden model takes the decision. Pages without text can be read with OCR that also runs in your browser.

How to split a PDF by content

  1. Add the PDF. The analysis starts at once and reads every page locally.
  2. Choose the detection sensitivity. Conservative proposes only very strong boundaries, Balanced is the default and suits mixed bundles, and Exploratory suggests more possible cuts. Changing it updates the proposals without reading the document again.
  3. Leave Analyze scanned pages with local OCR switched on if the PDF contains scans. Switching it off or on reads the document again.
  4. Review the list of pages. Each page shows a label, the reasons why it may start a new document and a confidence percentage. A checked page starts a new PDF, and the first page always starts the first part.
  5. Check or uncheck any page to correct the cuts, then press the create button, which shows how many local parts you will get.
  6. Download the ZIP, with the suffix _smart_split and a timestamp and one named PDF per part. A single part downloads as a PDF.

What the analysis looks for

Each signal found on a page adds to its confidence, and the sensitivity sets how much confidence a page needs before it is proposed. Blank pages are never proposed as the start of a document, so a blank separator stays at the end of the previous part.

Names, limits and what the parts keep

Privacy: no upload and no AI service

The text is read by the PDF engine bundled with OléPDF, the OCR engine is served by OléPDF itself, and the parts are written by a local PDF library. All of it runs in your browser tab. The document is never uploaded, no page is sent to an AI provider and no third-party script is involved.

Typical uses

Frequently asked questions

What does Smart Split PDF cost?

The free plan includes one split document per day. The 7-Day Pass and Pro remove the daily limit and allow splitting several documents in a single run.

Does it use artificial intelligence?

No AI service is called and no model decides for you. The proposals come from rules that read identifiers, page numbers, blank pages, headings and keywords, and every reason is shown next to its page.

What if a proposed cut is wrong?

Uncheck the page and it stays inside the previous part. Check any other page to start a new part there. Nothing is created until you press the create button.

Does it work with scanned PDFs?

Yes. With the OCR option on, up to 50 pages without text are read locally in each document. Pages that remain without text are counted below the list so you can place their cuts yourself.

What happens when I split several PDFs at once?

With the 7-Day Pass or Pro, each PDF is analyzed on its own with the same sensitivity and OCR settings, and the proposed cuts are applied automatically, without the review step. You get one ZIP with a folder per document and a small manifest.

Is my document uploaded?

No. The analysis, the OCR and the split all run in your browser, and the PDF never reaches a server.

Organize and Compress

All tools

Photos rather than PDFs? Try OléPIC, our image tools.

Videos rather than PDFs? Try OléVIDEO, our video tools.