n8n-nodes-docupipe

This is an n8n community node for DocuPipe - an intelligent document processing platform. It lets you automate document uploads, data extraction, classification, and more directly from your n8n workflows.

What is DocuPipe? DocuPipe processes documents (PDFs, images, Word files) using AI to extract structured data, classify documents into categories, and split/merge multi-page files. You define schemas (what fields to extract) and classes (document categories) in the DocuPipe dashboard, then use this node to automate the processing.

n8n is a fair-code licensed workflow automation platform.

Installation |
Credentials |
Operations |
Triggers |
Examples |
Compatibility |
Resources

Installation

Follow the installation guide in the n8n community nodes documentation.

Credentials

You need a DocuPipe API key to use this node:

1. Sign up for free or log in at app.docupipe.ai
2. Go to Settings → General
3. Copy your API Key
4. In n8n, create a new DocuPipe API credential and paste your key
5. Click Test to verify the connection

Operations

Document

• Upload - Upload a document from a URL, binary file (e.g. email attachment), or base64 string
• Get - Retrieve a document's details by its ID
• Split - Split a multi-page document into individual documents
• Merge - Merge multiple documents into one

Extraction

• Extract - Extract structured data from a document using a schema you defined in DocuPipe. Supports both V3 (default, latest agentic engine) and V2 (previous engine) via the API Version field.
• Get Result - Retrieve the extracted data after extraction completes
• Upload and Extract - Upload a document and extract data in one step. Supports both V3 (default) and V2 engines.

Classification

• Classify - Classify a document into categories you defined in DocuPipe

Schema / Class

• List Schemas - List all extraction schemas in your account
• List Classes - List all document classes in your account

Triggers

The DocuPipe Trigger node automatically starts a workflow when DocuPipe finishes processing. No polling needed - DocuPipe sends data to n8n in real-time via webhooks.

Examples

Example 1: Upload a file from URL and extract data

The simplest workflow - upload a document and extract structured data from it.

1. Add a Manual Trigger node (click to run)
2. Add a DocuPipe node → Resource: Extraction → Operation: Upload and Extract
3. Set Input Mode to URL
4. Enter the file URL (e.g. https://example.com/invoice.pdf) and filename
5. Select a Schema from the dropdown (create one first at app.docupipe.ai → Schemas)
6. Run the workflow - you'll get back a documentId and standardizationId
7. To get the extracted data, add a DocuPipe Trigger (Extraction Complete) in a separate workflow

Example 2: Process email attachments automatically

Automatically extract data from every email attachment you receive.

1. Add a Gmail Trigger node → Event: Message Received
2. Add a DocuPipe node → Resource: Document → Operation: Upload
3. Set Input Mode to Binary File
4. Leave Binary Property as data (Gmail outputs attachments as "data")
5. Connect Gmail → DocuPipe

Now every email attachment is automatically uploaded to DocuPipe. Add a DocuPipe Trigger (Document Processed) in another workflow to process them after upload.

Example 3: Extract data from Google Drive files

Process documents saved in Google Drive.

1. Add a Google Drive Trigger node → Event: File Created
2. Add a Google Drive node → Operation: Download (to get the file as binary)
3. Add a DocuPipe node → Resource: Extraction → Operation: Upload and Extract
4. Set Input Mode to Binary File
5. Leave Binary Property as data
6. Select your Schema
7. Connect: Google Drive Trigger → Google Drive Download → DocuPipe

Example 4: Extract data and send results to Google Sheets

End-to-end automation: extract invoice data and write it to a spreadsheet.

Workflow 1 - Kick off extraction:

1. Gmail Trigger (Message Received)
2. DocuPipe → Upload and Extract (Binary File mode, select your invoice schema)

Workflow 2 - Process results:

1. DocuPipe Trigger → Event: Extraction Complete (receives standardizationId and extracted data)
2. Google Sheets → Append Row (map the extracted fields to spreadsheet columns)

Example 5: Classify and route documents

Automatically classify documents and take different actions based on the type.

Workflow 1 - Upload and classify:

1. HTTP Request node → Download a file (set Response Format to File)
2. DocuPipe → Upload (Binary File mode)
3. DocuPipe → Classify (use the documentId from upload, select your classes)

Workflow 2 - Route based on classification:

1. DocuPipe Trigger → Event: Classification Complete (returns documentId and classIds)
2. IF node → Check which class was assigned
3. Route to different actions (e.g. invoices → accounting software, contracts → legal team)

Example 6: Upload a document from base64 (database or API)

If your documents are stored as base64 strings (e.g. in a database or returned by an API):

1. Add a node that outputs base64 content (e.g. a Postgres node querying a documents table, or a Code node)
2. Add a DocuPipe node → Resource: Document → Operation: Upload
3. Set Input Mode to Base64
4. Set Base64 Content to reference the field from the previous node (e.g. {{ $json.fileContent }})
5. Set Filename (e.g. invoice.pdf) - DocuPipe needs this to know the file format

Example 7: Split a document and process each part

Split a multi-page PDF into individual documents, then extract data from each one.

Workflow 1 - Upload and split:

1. Manual Trigger
2. DocuPipe → Upload (URL or Binary File)
3. DocuPipe → Split (use the documentId from upload)

Workflow 2 - Process each child document:

1. DocuPipe Trigger → Event: Split Complete (returns parentDocumentId and childDocumentIds)
2. Split In Batches node → loop through each childDocumentId
3. DocuPipe → Extract (use each childDocumentId + select a schema)

How Binary File mode works

Many n8n nodes output files as "binary data" (Gmail attachments, HTTP downloads, Google Drive files, Dropbox, etc.). DocuPipe's Binary File mode reads these files automatically - no manual conversion needed.

Steps:

1. Add a node that outputs a file (e.g. HTTP Request with Response Format: File)
2. Connect it to a DocuPipe node
3. Set Input Mode to Binary File
4. Leave Binary Property as data (this is the default for most nodes)
5. Run it - DocuPipe automatically reads the file and uploads it

How to check the binary property name: If "data" doesn't work, run the previous node first and click on the Binary tab in the output. The property name is shown there (e.g. "data", "attachment_0", "file").

AI Agent Support

This node supports usableAsTool, making it available as a tool in n8n's AI Agent workflows. Self-hosted users need to set the environment variable:

N8N_COMMUNITY_PACKAGES_ALLOW_TOOL_USAGE=true

Compatibility

• Tested with n8n version 2.12.3
• Requires Node.js 18 or later

Upgrading from 0.1.2 to 0.1.3

The Extract and Upload+Extract operations now support both V3 (default) and V2 engines, with parameters reorganized into version-specific Additional Fields and Advanced Options collections.

For existing workflows, please re-open each Extract / Upload+Extract node and:

1. API Version defaults to V3. If you want to keep using the V2 engine, switch it to V2.
2. Re-add any settings you previously configured under Additional Fields (Display Mode, Split Mode, Effort Level, Guidelines, Dataset). These do not auto-migrate because the parameter was renamed.
3. Pages, Timeout, Metadata, and Use Metadata are now under Advanced Options.

Workflows that used Extract with no Additional Fields will keep working but will route to the V3 engine, which may produce slightly different extractions than V2. Switch to V2 if you need byte-identical output.

Resources

• DocuPipe Dashboard - Create schemas, classes, and manage documents
• DocuPipe Documentation - API reference and guides
• DocuPipe Website - Sign up for a free account
• n8n Community Nodes Documentation