Agent framework integrations

See also the TypeScript integrations overview for the equivalent guides in the TypeScript SDK.

Neo4j Agent Memory provides official integrations for eleven popular agent frameworks and cloud providers, enabling persistent memory storage backed by Neo4j’s graph database.

Available integrations

Framework Use Case

Framework	Use Case
Use with LangChain	Chains and agents with conversation memory
Use with LlamaIndex	RAG applications with document + graph retrieval
Use with PydanticAI	Modern type-safe agents with automatic tracing
Use with CrewAI	Multi-agent systems with shared memory
Use with the OpenAI Agents SDK	OpenAI function calling with persistent memory
Use Vertex AI embeddings with Neo4j	Vertex AI embeddings for a Neo4j-backed client, with links to the ADK and MCP guides
Use with Google ADK	Memory service for ADK `Runner` agents
Use with Strands agents	AWS Strands SDK with Bedrock and Context Graph tools
Use Amazon Bedrock embeddings	Titan and Cohere embeddings via Bedrock
Route memory searches with the hybrid provider	Route message, entity and preference searches inside one Neo4j-backed client
Use with Microsoft Agent Framework	Context providers, GDS algorithms, and graph-enhanced memory

Use with LangChain

Chains and agents with conversation memory

Use with LlamaIndex

RAG applications with document + graph retrieval

Use with PydanticAI

Modern type-safe agents with automatic tracing

Use with CrewAI

Multi-agent systems with shared memory

Use with the OpenAI Agents SDK

OpenAI function calling with persistent memory

Use Vertex AI embeddings with Neo4j

Vertex AI embeddings for a Neo4j-backed client, with links to the ADK and MCP guides