​
Search Guide

Cognee provides powerful search capabilities to query your knowledge graph. This guide explains the different search types and how to use them effectively.
Note: For a comprehensive reference of all available search types, see the Search Types Reference.

​
Search Types

Cognee offers several main types of search: Insights search helps you discover connections and relationships in your knowledge graph: 
from cognee.modules.search.types import SearchType
import cognee
import asyncio

async def main():
    results = await cognee.search(
        query_text="What is the relationship between Einstein and Germany?",
        query_type=SearchType.INSIGHTS
    )
    
    for result in results:
        print(result)

if __name__ == "__main__":
    asyncio.run(main())
Use insights search when you want to:
  • Discover relationships between entities
  • Understand how concepts are connected
  • Get a deeper understanding of your knowledge graph
RAG search combines retrieval from your knowledge graph with generative AI to provide comprehensive answers: 
results = await cognee.search(
    query_text="Explain Einstein's contributions to physics",
    query_type=SearchType.RAG_COMPLETION
)
Use RAG search when you want to:
  • Get detailed explanations
  • Combine multiple pieces of information
  • Generate comprehensive answers based on your knowledge graph
Get contextually aware answers using graph traversal 
results = await cognee.search(
    query_text="List all the important characters in Alice in Wonderland.",
    query_type=SearchType.GRAPH_COMPLETION
)

​
4. Additional Search Types

Cognee offers several additional specialized search types:
  • SUMMARIES: Get concise summaries of topics
  • CODE: Find code-related information
  • GRAPH_SUMMARY_COMPLETION: Combine graph traversal with summarization
  • GRAPH_COMPLETION_COT: Combine Chain of Thought with graph traversal
  • GRAPH_COMPLETION_CONTEXT_EXTENSION: Extend context iteratively with graph traversal
  • CYPHER: Execute raw Cypher queries directly against your graph database
For details on these and other search types, see the Search Types Reference.

​
Advanced Search Options

​
Filtering by Node Type

You can filter search results by node type: 
results = await cognee.search(
    query_text="Your query",
    query_type=SearchType.INSIGHTS,
    node_types=["Person", "Organization"]  # Only include these node types
)

​
Filtering subgraph by Node Type and Node Names

You can filter search results in graph completion searches by node type and node names: 
from cognee.modules.engine.models import NodeSet

results = await cognee.search(
    query_text="Your query",
    query_type=SearchType.GRAPH_COMPLETION,
    node_type=NodeSet,
    node_name=['my_nodeset_name_1', 'my_nodeset_name_2']
)
This functionality works with the following searches (only Neo4j and Kuzu):
  • GRAPH_COMPLETION
  • GRAPH_SUMMARY_COMPLETION
  • GRAPH_COMPLETION_COT
  • GRAPH_COMPLETION_CONTEXT_EXTENSION

​
Searching with Context

You can provide additional context for your search: 
results = await cognee.search(
    query_text="Your query",
    query_type=SearchType.INSIGHTS,
    context="Additional context to help refine the search"
)

​
Search Examples

​
Example 1: Finding Connections Between People

# Add some content
await cognee.add("Steve Jobs co-founded Apple with Steve Wozniak in 1976.")
await cognee.add("Tim Cook became CEO of Apple after Steve Jobs.")

# Process the content
await cognee.cognify()

# Search for connections
results = await cognee.search(
    query_text="What is the relationship between Steve Jobs and Tim Cook?",
    query_type=SearchType.INSIGHTS
)

​
Example 2: Generating Comprehensive Answers

# Add some content
await cognee.add("Machine learning is a subset of artificial intelligence.")
await cognee.add("Deep learning is a type of machine learning that uses neural networks.")
await cognee.add("Neural networks are computing systems inspired by biological neural networks.")

# Process the content
await cognee.cognify()

# Generate a comprehensive answer
results = await cognee.search(
    query_text="Explain the relationship between machine learning, deep learning, and neural networks",
    query_type=SearchType.RAG_COMPLETION
)
If you’re not getting the expected search results:
  1. No results: Make sure you’ve added relevant content and run cognee.cognify()
  2. Irrelevant results: Try refining your query or using a different search type
  3. Missing connections: Your knowledge graph might not have captured the relationships you’re looking for; try adding more relevant content

​
Next Steps