PARSER GUIDES

Split Documents

ParsersParser Settings

Split large PDFs into separate documents

The Split Documents feature lets you break a large PDF into smaller documents before PARSER reads it. Each split document appears as its own document in your inbox, making it easier to review, approve, and export.

Document classifiers have a similar setting called Split Before Routing. Use that when a PDF may contain several document types that should be routed to different parsers.

When to use split documents

Use this feature when:

• You have a large PDF with multiple documents (e.g. a PDF containing 10 separate invoices)
• You want to process each document separately rather than as one big file
• You want to group one long PDF by a changing value, such as statement date, service date, customer name, claim ID, policy number, or invoice number
• You need to organize or review documents individually

How to enable split documents

1. Go to your parser's Settings -> Parser Settings.

2. Find Split PDF files into smaller documents.

3. Turn on the toggle.

4. Choose a split method.

Note: You can't use Split Documents and Process email attachments together (in Email settings) at the same time. If you enable one, the other is automatically disabled.

Split methods

Once enabled, you can choose how to split your PDFs:

Split every X pages

Split your document into multiple files, with each file containing the same number of pages.

Example: If you set it to 5 pages, a 20-page PDF becomes 4 separate documents with 5 pages each.

How to use:

1. Select Split every X pages.

2. Enter the number of pages per document (e.g. 5).

3. Upload your PDF. It will be split automatically.

Split by page ranges

Enter the page numbers where you want to split the PDF. Each number is the last page of that document (a split happens after that page). Use commas to separate numbers and a hyphen for a range (e.g. 7-9 means 7, 8, and 9).

How to use:

1. Select Split by page ranges.

2. Enter the split points: the last page of each resulting document, separated by commas. Ranges like 5-10 mean split after pages 5, 6, 7, 8, 9, and 10.

Example: 1, 3, 7-9 on a 10-page PDF

This splits after pages 1, 3, 7, 8, and 9, giving 6 documents:

Document

Pages

#1

Page 1

#2

Pages 2-3

#3

Pages 4-7

#4

Page 8

#5

Page 9

#6

Page 10

More examples:

• 5 -> Two documents: pages 1-5 and pages 6 to the end
• 1, 5, 10 -> Four documents: page 1; pages 2-5; pages 6-10; pages 11 to the end
• 2, 4, 6 on a 6-page PDF -> Three documents: pages 1-2, 3-4, 5-6

Split with AI

Use Split with AI when the number of pages changes from document to document, or when you want PARSER to split by information inside the PDF.

After selecting Split with AI, choose how AI should split the PDF:

• When a new document starts - Best for PDFs that contain multiple invoices, statements, forms, reports, applications, or records mixed together. PARSER looks for where one document ends and the next one begins.
• When a field value changes - Best when one long PDF should be grouped by a value like date, customer, invoice number, claim ID, policy number, or a table value such as transaction date.

For When a field value changes, choose a field from your parser or enter a custom value. Field choices can include regular fields and values inside tables, such as a transaction date in a transactions table.

You can also add an optional hint, for example:

• Each invoice starts with an invoice number at the top.
• Use service date, not due date.
• Group transactions by transaction date.

Availability and cost: Split with AI is available on Pro, Scale, and higher plans. It costs an extra 1 credit per page being split. If Pages to be parsed is enabled, this charge applies only to the selected pages.

How it works

1. Select pages: If the parser has Pages to be parsed enabled, PARSER first keeps only the selected pages. Otherwise, it uses the full PDF.

2. Split the selected pages: PARSER applies the fixed or AI split setting to those pages.

3. Process each part: Each split document is processed separately using your parser's fields and settings. For classifiers, each split document is routed separately before the target parser extracts data.

4. Results: You'll see multiple documents in your Document Inbox, one for each split.

For example, if Pages to be parsed is 1-6 on a 1,000-page PDF, PARSER splits and extracts only pages 1 through 6. It does not split, extract, or apply per-page charges to the other 994 pages.

Parser split settings apply to PDFs added through the app, email, upload pages, embedded upload forms, API imports, URL imports, and Zapier. Classifier split settings apply to classifier uploads, classifier email intake, API classifier uploads, URL imports, and Zapier imports.

Important notes

• PDFs only: Split Documents works with PDF files only. Other file types are not split.
• Processing time: Splitting adds time before the document is processed. Split with AI can take longer than fixed page splitting because PARSER checks the document content first.
• Credits: Each split document uses credits based on its page count. A 20-page PDF split into 4 documents of 5 pages each uses the same standard processing credits as the 20 selected pages. Split with AI costs an extra 1 credit per selected page.
• File names: Each split document is named from the original file and the split (e.g. with a page or part indicator). The exact format depends on the split method.

Use cases

Monthly statements

If you receive a monthly statement PDF and need one document per date, use Split with AI -> When a field value changes and choose the date field you want to group by, such as transaction date.

Batch invoices

If a vendor sends one PDF with 20 invoices, use Split with AI -> When a new document starts if invoice length varies. If every invoice has the same page count, Split every X pages is faster and simpler.

Multi-document PDFs

If you have a PDF containing different types of documents, use Split with AI -> When a new document starts so PARSER can separate each document based on its content.

If those split documents should go to different parsers, use Document Classifiers and enable Split Before Routing in classifier settings.

Troubleshooting

Split not working:

• Ensure you're uploading a PDF file (not an image or other format)
• Check that Split Documents is enabled in Parser Settings
• Verify your split settings are correct (e.g. page numbers are valid)

Too many or too few splits:

• Adjust your Split every X pages number
• Review your page ranges to ensure they match your document structure
• For Split with AI, add a clearer hint or choose a more specific field value
• For value-based AI splitting, make sure the selected field exists in the document and represents the grouping you want

Next steps

• Pages to be parsed - Limit which PDF pages are extracted
• Parser Settings -> Email Processing - Process documents via email
• Document Classifiers - Split and route mixed document batches
• Uploading Documents - Learn more about uploading
• How to Split Documents - FAQ on splitting documents