n8n LlamaCloud Nodes

This repository contains custom n8n nodes for integrating with LlamaCloud, providing powerful document processing and retrieval capabilities within your n8n workflows.

🚀 Features

This package includes three custom nodes:

📄 LlamaParse

• Parse PDF files and extract their content in markdown format
• Uses LlamaCloud's document parsing capabilities
• Perfect for document preprocessing workflows

🔍 LlamaExtract

• Extract structured data from files using LlamaCloud extraction agents
• Get elegant, structured information from documents
• Ideal for data extraction and analysis workflows

💬 LlamaCloud

• Retrieve context from your LlamaCloud indexes
• Chat with your indexed documents
• Great for building RAG (Retrieval-Augmented Generation) applications

📋 Prerequisites

Before using these nodes, you need:

1. Node.js and npm (Minimum version Node 20)

   You can find instructions on how to install both using
   nvm (Node Version Manager) for Linux, Mac, and WSL here. For
   Windows users, refer to Microsoft's guide to Install NodeJS on Windows.

2. n8n installed globally

   npm install n8n -g
   

3. LlamaCloud API Key

   • Sign up at cloud.llamaindex.ai
   • Get your API key from the dashboard

🛠️ Installation

Option 1: Local Development (Recommended)

1. Clone this repository

   git clone https://github.com/run-llama/n8n-llamacloud.git
   cd n8n-llamacloud
   

2. Install dependencies

   npm install
   

3. Build the nodes and publish locally

   npm run build
   npm link
   

4. Link to n8n custom nodes directory

   # Create custom nodes directory if it doesn't exist
   mkdir -p ~/.n8n/custom
   npm link n8n-llamacloud
   

5. Restart n8n

   n8n stop
   n8n start
   

Option 2: Global Installation

1. Install globally

   npm install -g n8n-llamacloud
   npm link
   

2. Restart n8n

   n8n stop
   n8n start
   

🔧 Setup

1. Configure LlamaCloud Credentials

1. Open n8n in your browser (usually http://localhost:5678)
2. Go to Settings → Credentials
3. Click Add Credential
4. Search for "LlamaCloud API Key"
5. Enter your LlamaCloud API key
6. Test the connection and save

2. Create LlamaCloud Indexes (for LlamaCloud node)

1. Go to cloud.llamaindex.ai
2. Create a new project
3. Upload documents to create an index
4. Note the index name for use in the LlamaCloud node

📚 Resources

• LlamaCloud Documentation
• LlamaIndex Documentation
• n8n Documentation
• n8n Community

📖 Usage Examples

Example 1: Document Processing Pipeline

graph LR
    A[PDF File] --> B[LlamaParse]
    B --> C[LlamaExtract]
    C --> D[Structured Data]

Workflow:

1. LlamaParse node: Parse a PDF file to markdown
2. LlamaExtract node: Extract structured data using an extraction agent
3. Use the structured data in subsequent nodes

Example 2: RAG Chat Application

graph LR
    A[User Query] --> B[LlamaCloud]
    B --> C[Retrieved Context]
    C --> D[AI Response]

Workflow:

1. HTTP Request node: Receive user query
2. LlamaCloud node: Retrieve relevant context from your index
3. OpenAI node: Generate response using the retrieved context

Example 3: Document Analysis

graph LR
    A[Document] --> B[LlamaParse]
    B --> C[LlamaExtract]
    C --> D[Analysis Results]

Workflow:

1. LlamaParse node: Convert document to markdown
2. LlamaExtract node: Extract specific information (e.g., dates, amounts, entities)
3. Code node: Process and analyze the extracted data

🔍 Node Details

LlamaParse Node

Purpose: Parse PDF files and extract content in markdown format

Parameters:

• File Path (required): Path to the PDF file to parse
  • Example: /Users/username/Documents/document.pdf

Output: Markdown content of the parsed document

LlamaExtract Node

Purpose: Extract structured data from files using LlamaCloud extraction agents

Parameters:

• Agent ID (required): The ID of your LlamaCloud extraction agent
• File Path (required): Path to the file to extract data from

Output: Structured data based on your extraction agent's configuration

LlamaCloud Node

Purpose: Retrieve context from your LlamaCloud indexes

Parameters:

• Index Name (required): Name of your LlamaCloud index
• Chat Input (from previous node): The query to search for in your index

Output: Retrieved context from your index