Graph RAG nodes are advanced knowledge graph-powered RAG components that combine document processing, entity extraction, relationship mapping, and intelligent retrieval into a single unit.

​
What are Graph RAG Nodes?

Graph RAG nodes automatically handle:
  • Document Processing: Intelligent document ingestion and preparation
  • Entity Extraction: Advanced entity recognition using large language models
  • Relationship Mapping: Automated identification of entity relationships
  • Knowledge Graph Construction: Building and maintaining knowledge graphs
  • Semantic Retrieval: Query-based document retrieval enhanced with graph context

​
Key Benefits

  • Knowledge Graph Intelligence: Leverages extracted entities and relationships for deeper understanding
  • Semantic Context: Provides rich contextual information beyond traditional keyword matching
  • Relationship Discovery: Automatically identifies connections between concepts and entities
  • Scalable Knowledge: Efficiently handles large-scale knowledge graph construction and querying

​
Available Endpoints

EndpointMethodPurposeDocumentation
List Graph RAG NodesGETRetrieve all graph RAG nodes from a flow📄 List Documentation
Update Graph RAG ConfigurationPATCHModify graph RAG node settings🔧 Update Documentation

​
Base URL Structure

https://{flow_name}.flows.graphorlm.com/graph-rag

​
Configuration

Graph RAG nodes have a simple configuration approach:
ParameterTypeDefaultDescription
topKinteger | null5Number of top results to retrieve. Set to null for unlimited processing

​
Example Configuration

{
  "id": "graph-rag-1748287628685",
  "type": "graph-rag",
  "data": {
    "name": "Knowledge Graph RAG",
    "config": {
      "topK": 15
    },
    "result": {
      "updated": true,
      "processing": false,
      "total_processed": 2150,
      "total_chunks": 640,
      "total_entities": 1280,
      "total_relationships": 850
    }
  }
}

​
Strategy Selection

​
Precision Strategy

Configuration: topK: 5-10
  • Fast, resource-efficient
  • High-relevance entity-based retrieval
  • Best for expert systems and real-time applications

​
Balanced Strategy

Configuration: topK: 12-20
  • Good overall performance
  • Balanced entity coverage and relationship exploration
  • Best for general-purpose knowledge management

​
Comprehensive Strategy

Configuration: topK: 25-40
  • Thorough knowledge graph exploration
  • High entity recall with comprehensive relationship coverage
  • Best for research platforms and complex domains

​
Unlimited Strategy

Configuration: topK: null
  • Complete knowledge graph analysis
  • Maximum entity and relationship coverage
  • Best for exhaustive knowledge discovery

​
Authentication

All endpoints require API token authentication: 
Authorization: Bearer YOUR_API_TOKEN
Generate API tokens through the API Tokens guide.

​
Response Formats

​
Success Response

{
  "success": true,
  "message": "Operation completed successfully",
  "node_id": "graph-rag-1748287628685"
}

​
Error Response

{
  "detail": "Descriptive error message"
}

​
Basic Usage Example

import requests

# Configuration
flow_name = "my-rag-pipeline"
api_token = "YOUR_API_TOKEN"
base_url = f"https://{flow_name}.flows.graphorlm.com"
headers = {"Authorization": f"Bearer {api_token}"}

# Get all graph RAG nodes
response = requests.get(f"{base_url}/graph-rag", headers=headers)
nodes = response.json()

# Update a node configuration
node_id = "graph-rag-1748287628685"
update_data = {"config": {"topK": 20}}

response = requests.patch(
    f"{base_url}/graph-rag/{node_id}",
    headers={**headers, "Content-Type": "application/json"},
    json=update_data
)

​
Best Practices

  • Start with Balanced Strategy: Use topK: 15 for most applications
  • Monitor Performance: Track entity extraction quality and processing time
  • Regular Reviews: Periodically review and optimize configurations
  • Quality First: Prioritize entity and relationship quality over quantity

​
Next Steps

List Graph RAG Nodes

Learn how to retrieve graph RAG node configurations

Update Graph RAG Configuration

Master graph RAG configuration management

Dataset Management

Understand dataset integration with graph RAG nodes

Flow Orchestration

Learn about comprehensive flow management