Skip to Content
ReferenceAPI

Cognee API Reference

Overview

The Cognee API provides a comprehensive set of functions and endpoints for building, managing, and querying knowledge graphs. This reference outlines both the Python API functions and the HTTP REST API endpoints.

Core Python API Functions

Data Management

# Add content to Cognee
await cognee.add(content)
ParameterTypeDescription
contentstrText content, file path (using file:// prefix), or URL to add to the knowledge graph
# Process content and build the knowledge graph
await cognee.cognify()
# Reset data and system state
await cognee.prune.prune_data()  # Reset just the data
await cognee.prune.prune_system(metadata=True)  # Reset system state
await cognee.prune.prune_all()  # Reset everything

Search and Query

# Search the knowledge graph
from cognee.modules.search.types import SearchType
 
results = await cognee.search(
    query_text="Your search query",
    query_type=SearchType.INSIGHTS,  # or INFORMATION, RAG, etc.
)
ParameterTypeDescription
query_textstrThe search query text
query_typeSearchTypeType of search to perform (see Search Types Reference for all options)

Note: For a comprehensive reference of all available search types, see the Search Types Reference.

Graph Visualization

import cognee.api.v1 as api
 
# Generate an HTML visualization of your graph
visualization_path = await api.render_graph()

The visualization is saved as an HTML file in the .cognee_system folder.

Configuration

# Configure Cognee
import cognee
 
# LLM configuration
cognee.config.llm_api_key = "YOUR_API_KEY"
cognee.config.llm_model = "gpt-4"  # Default model
 
# Database configuration
cognee.config.db_provider = "sqlite"  # or "postgres"
cognee.config.db_connection_string = "postgresql://user:password@localhost:5432/cognee"

HTTP REST API

The Cognee HTTP API provides RESTful endpoints for interacting with your knowledge graph from any application.

Interactive API Documentation

Below is the interactive API documentation generated from our OpenAPI specification:

Key Endpoints

  • POST /api/v1/add: Add content to the knowledge graph
  • POST /api/v1/cognify: Process content and build the knowledge graph
  • POST /api/v1/search: Search the knowledge graph
  • GET /api/v1/render_graph: Generate a visualization of the knowledge graph
  • POST /api/v1/prune: Reset data and system state

Using the API in Your Applications

Python Example

import cognee
import asyncio
 
async def main():
    # Configure
    cognee.config.llm_api_key = "YOUR_API_KEY"
    
    # Add content
    await cognee.add("Your content here")
    
    # Process
    await cognee.cognify()
    
    # Search
    from cognee.modules.search.types import SearchType
    results = await cognee.search("Your query", query_type=SearchType.INSIGHTS)
    
    # Visualize
    import cognee.api.v1 as api
    visualization_path = await api.render_graph()
    
if __name__ == "__main__":
    asyncio.run(main())

HTTP API Example

import requests
 
# Add content
requests.post(
    "http://localhost:8000/api/v1/add",
    json={"content": "Your content here"}
)
 
# Process
requests.post("http://localhost:8000/api/v1/cognify")
 
# Search
response = requests.post(
    "http://localhost:8000/api/v1/search",
    json={
        "query_text": "Your query",
        "query_type": "INSIGHTS",
    }
)
results = response.json()

Understanding Graph Visualization

The graph visualization shows:

  • Nodes: Entities extracted from your content (people, places, concepts, etc.)
  • Edges: Relationships between entities
  • Categories: Nodes are colored by category (e.g., person, location, organization)

The visualization helps you understand how different pieces of information are connected in your knowledge graph.

Next Steps

SDKSearch Types