This article explains how to split multi-page documents supported by Parseur — including PDFs and TIFF images — into individual documents.

Why Split Multi-Page Documents?

Sometimes, your PDF or TIFF image contains multiple documents bundled into a single file. While you can create a template or use Parseur’s AI engine to extract data from the first document, Parseur requires splitting the original file into individual documents to extract data from each section properly.

Using the Split Document feature ensures that all documents within the bundle are processed, as if looping through each one separately.

Setting Up the Split Document Feature

Parseur allows you to automate this process by specifying how many pages each split should contain. Here’s how to configure the settings:

Open your Parseur mailbox.
Click on Upload / Import from the left menu.
Click on Adjust split settings under the drag-and-drop upload box:

There are four controls around document splitting:

Split documents every X amount of pages
- Self explanatory, Parseur will split the document by a static amount of pages.
Split documents by page ranges
- Many users have large documents including only a few pages of data they need to process. Contrasting from our keep pages by range function, this option allows you to split a document into multiple documents from the page ranges you define.
Split documents by keyword
- You can split documents by keywords, telling it to split before or after every instance of a keyword.
Split documents using AI
- Lastly, you can use AI to split bundled PDFs into separate documents, even when they have different page counts and no clear start/end keywords. Examples on when to use this are below.

Frequently Asked Questions

Why is Parseur only processing the first page of my document?

If each page of your multi-page document contains repeating data, use the Split Document feature to ensure that Parseur processes all pages individually. After enabling this feature, delete the original upload and re-upload the document to apply the new settings.

What’s the difference between Split Document and Table Fields?

Both features capture repetitive patterns, but they serve different purposes:

Split Document: Use this feature when the entire document contains identical fields repeated across multiple pages or sections. It will divide the original file into separate documents, each processed individually.
Table Fields: Use this feature to extract data organized in tables or lists within a single page or across multiple pages.

Can I use different splitting settings for different document types?

The split settings are global for each mailbox. If you need different split configurations, you can either:

Adjust the split settings manually each time you upload documents, or
Create separate mailboxes for each document type with unique split settings.

Can I split a document I've already uploaded?

Yes, when page splitting is enabled on your mailbox, you can manually split a document by opening it up in the Documents tab and clicking the Split button:

Can I reverse the document splitting?

After your documents have been processed, you will see the original document uploaded with a Split status, when clicking on this document you are given a few options to either reverse the document split or to re-split the document again:

Note: The Reverse document split and re-split document functions will consume credits on use.

Tip: If you are working with a single PDF or TIFF that contains multiple documents, each with their own tables, you can combine Split Document and Table Fields.

When should I use the AI for document splitting?

We plan to eventually make document splitting with AI enabled by default for users given its many advantages:

Scenario	Why Keyword / Page Split Fails	When Keyword /Page Split Works	Why AI Splitting is Better
Irregular Document Lengths (e.g., invoices of 1, 3, or 7 pages in one PDF)	Fixed page counts will over/under split, since document lengths vary.	Works only if every document is always the same number of pages.	AI can detect document boundaries using headers, footers, or visual patterns, and handle varying lengths automatically.
Keywords Repeated Inside Documents (e.g., "Invoice Total" appears mid-document)	Splitting by keyword may cut documents in the middle.	Works well if keyword is unique and always at the start of a document.	AI can understand context (e.g., “Invoice Total” in a footer ≠ start of a new document) and avoid false splits.
Mixed Document Types (e.g., invoices, POs, credit memos bundled together)	Keyword-based split may confuse document types if they share similar terms (e.g., "Invoice" in a PO).	Works only if each type has a unique, consistent keyword that always appears on first page.	AI can classify page type, group similar pages together, and output separate documents per type.
Multi-Page Continuations (e.g., invoice spanning several pages without repeating keyword)	Second/third pages have no start keyword → they may be separated incorrectly.	Works if each page starts with the keyword or if page ranges are always predictable.	AI can link continuation pages using repeated invoice number or layout similarity, keeping them together.
Visual-Only Boundaries (e.g., logo, barcode, blank page)	No keyword to match → can’t split correctly.	Doesn't work at all without text clues.	AI can use layout, images, logos, or barcodes to identify document start pages.
Unstructured / Low-Quality PDFs (e.g., scanned docs with OCR errors)	Keywords may be missed entirely due to OCR mistakes.	Works only if text is always machine-readable and reliable.	AI can use pattern recognition and fuzzy matching to still detect likely boundaries even with imperfect text.