langchain-cognee

This package contains the LangChain integration with cognee.

This package enables you to:

• Ingest documents into cognee
• Build or update a knowledge graph
• Retrieve and query your data using LangChain's standard interfaces

For more information, check out cognee documentation.

Installation

pip install -U langchain-cognee

Configuration

Set your environment variables required by cognee:

export LLM_API_KEY="your-openai-api-key"

Cognee's default settings:

• LLM Provider: OpenAI
• Databases: SQLite, LanceDB, networkx

In case you want to customize your settings, please refer here and configure your env variables accordingly.

Supported databases

• Relational databases: SQLite, PostgreSQL
• Vector databases: LanceDB, PGVector, QDrant, Weviate
• Graph databases: Neo4j, NetworkX

Basic Usage

Below is a minimal example of how to use this integration:

    from langchain_cognee.retrievers import CogneeRetriever
    from langchain_core.documents import Document

    # 1) Instantiate the retriever
    retriever = CogneeRetriever(
        llm_api_key="YOUR_KEY", 
        dataset_name="test_dataset", 
        k=3
    )

    # 2) (Optional) Reset dataset if you want a clean slate
    retriever.reset_dataset()

    # 3) Add documents
    docs = [
        Document(page_content="Elon Musk is the CEO of SpaceX."),
        Document(page_content="SpaceX focuses on rockets."),
    ]
    retriever.add_documents(docs)

    # 4) Build knowledge graph
    retriever.process_data()

    # 5) Retrieve documents
    results = retriever.invoke("Tell me about Elon Musk")
    for doc in results:
        print(doc.page_content)

You can also incorporate CogneeRetriever in any LangChain chain.