Split Multi-Document PDFs with AI & pdfRest in n8n

Split multi-document PDFs into clean sections with AI—inside n8n

This n8n automation lets you upload a single PDF that contains multiple documents, uses pdfRest plus AI to suggest how the pages should be grouped, and then outputs separated PDFs and a downloadable ZIP with a page manifest.

What this workflow does

• Validates input: receives a PDF via an n8n upload form and verifies that exactly one valid PDF was provided.
• Prepares the PDF for analysis: uses pdfRest to inspect the PDF, run OCR to create a searchable master copy, and split it into individual page PDFs.
• Builds a page inventory: converts each page to Markdown, compiles a full-page inventory with original page numbers, and flags pages with low text.
• Uses AI to propose boundaries: sends the combined extracted text to OpenAI using the n8n LangChain Information Extractor to propose contiguous document groups with document type, title, counterparty, and evidence-based reasoning.
• Guided reviewer approval: presents editable titles, categories, and start/end page ranges for each suggested document; reviewers can exclude groups or add up to two additional groups, then approve or cancel.
• Ensures page correctness: checks that every page belongs to exactly one included group and returns clear correction messages for invalid ranges (edits are preserved across up to three review submissions).
• Generates outputs: creates descriptive split PDFs, optionally uploads them to Google Drive, builds a page-to-document manifest, compresses everything into a ZIP, and returns it to the user.

Use cases

• Separate mixed contract packets (e.g., terms + amendments) submitted as one PDF.
• Group invoice/statement documents within scanned multi-page files using OCR + AI page boundaries.
• Extract a structured inventory for compliance workflows by preserving original page numbers.

Technical details

• Workflow category: n8n automation workflow
• Nodes/logic used: if, set, code, form, sticky note, compression
• Key services: pdfRest (OCR, Markdown extraction, page splitting), OpenAI via n8n LangChain Information Extractor, optional Google Drive upload