Flexible GraphRAG
Python, LlamaIndex, LangChain, Docker Compose: 15 Property Graph, 4 RDF , 10 Vector, OpenSearch, Elasticsearch, Alfresco DBs. 13 data sources (9 auto-sync), KG auto-building, Ontologies, LLMs, Docling or LlamaParse doc processing, GraphRAG, RAG only,
What is Flexible GraphRAG?
Flexible GraphRAG is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to python, llamaindex, langchain, docker compose: 15 property graph, 4 rdf , 10 vector, opensearch, elasticsearch, alfresco dbs. 13 data sources (9 auto-sync), kg auto-building, ontologies, llms, docling...
Python, LlamaIndex, LangChain, Docker Compose: 15 Property Graph, 4 RDF , 10 Vector, OpenSearch, Elasticsearch, Alfresco DBs. 13 data sources (9 auto-sync), KG auto-building, Ontologies, LLMs, Docling or LlamaParse doc processing, GraphRAG, RAG only,
This server falls under the Data Science & ML and Knowledge & Memory categories on MCPgee, the world's largest MCP server directory with 33,000+ servers.
Features
- Python, LlamaIndex, LangChain, Docker Compose: 15 Property G
Use Cases
Maintainer
Works with
Installation
Manual Installation
npx flexible-graphragConfiguration
Configuration Details
claude_desktop_config.json
Performance
Response Metrics
Resource Usage
How to Set Up and Use Flexible GraphRAG
Flexible GraphRAG is a comprehensive Python-based RAG (Retrieval-Augmented Generation) and knowledge graph platform that combines 15 property graph databases, 4 RDF stores, 10 vector databases, and full-text search into a unified system for building AI knowledge applications. It supports 13 data sources (9 with auto-sync), automatic knowledge graph construction from documents, and exposes 9 MCP tools for ingestion, hybrid search, and Q&A. It works with multiple LLM providers (OpenAI, Anthropic, Google, Ollama, and more) and document processors (Docling and LlamaParse).
Prerequisites
- Python 3.13 recommended (3.10+ supported), with uv package manager
- Docker and Docker Compose for running graph and vector databases locally
- An API key for your chosen LLM provider (e.g., OPENAI_API_KEY for OpenAI, or Ollama running locally)
- At least one configured database (property graph, vector, or search backend)
- An MCP-compatible client such as Claude Desktop
Create a Python virtual environment and install
Use uv to create a Python 3.13 virtual environment and install the flexible-graphrag package from PyPI.
uv venv venv-3.13 --python 3.13
source venv-3.13/bin/activate
uv pip install flexible-graphragInstall the MCP server package
Install the separate MCP server package that exposes flexible-graphrag as MCP tools.
uv pip install flexible-graphrag-mcpConfigure environment variables
Set the core environment variables for your LLM provider and database backends. At minimum, set LLM_PROVIDER and the corresponding API key. Choose your VECTOR_DB, PG_GRAPH_DB, and SEARCH_DB based on what you have running.
export LLM_PROVIDER=openai
export OPENAI_API_KEY=your_openai_api_key
export OPENAI_MODEL=gpt-4o-mini
export VECTOR_DB=qdrant
export PG_GRAPH_DB=neo4j
export SEARCH_DB=BM25
export DOCUMENT_PARSER=doclingStart the backend services with Docker Compose
Use the provided Docker Compose file to start the graph and vector database services.
cd docker
docker-compose -f docker-compose.yaml -p flexible-graphrag up -dStart the MCP server
Launch the flexible-graphrag MCP server with HTTP transport on port 3001.
flexible-graphrag-mcp --http --port 3001Configure Claude Desktop to connect
Add the MCP server to your Claude Desktop configuration using the mcp-remote adapter to connect to the HTTP endpoint.
Ingest documents and start querying
Use the ingestion tools to load documents from files, URLs, or cloud storage. Then use the hybrid search and Q&A tools to query your knowledge base through Claude.
Flexible GraphRAG Examples
Client configuration
Connect Claude Desktop to the running flexible-graphrag MCP server via HTTP transport.
{
"mcpServers": {
"flexible-graphrag": {
"command": "npx",
"args": ["-y", "mcp-remote", "http://localhost:3001/mcp"],
"env": {
"OPENAI_API_KEY": "your_openai_api_key"
}
}
}
}Prompts to try
After ingesting documents, use these prompts with Claude to query your knowledge graph.
- "Ingest the PDF at /path/to/research-paper.pdf into the knowledge base"
- "Search for information about transformer architectures in my documents"
- "What are the key relationships between entities mentioned in the uploaded reports?"
- "Run a hybrid search combining graph and vector search for 'customer churn prediction'"
- "Answer: what were the main findings from the Q3 financial reports I uploaded?"Troubleshooting Flexible GraphRAG
Database connection errors on startup
Ensure Docker Compose services are running before starting the MCP server. Check that PG_GRAPH_DB and VECTOR_DB environment variables match a running database. Run 'docker-compose ps' to verify all services are healthy. The API health endpoint at http://localhost:8000/api/health shows connection status.
Document parsing fails or is very slow
Docling (the default parser) runs locally and may be slow on CPU. Set DOCLING_DEVICE=cuda or mps for GPU acceleration, or set DOCLING_OCR=false if OCR is not needed. For cloud-based parsing, set DOCUMENT_PARSER=llamaparse and provide LLAMAPARSE_API_KEY.
LLM provider authentication errors
Verify your API key is exported in the environment before starting the server. For OpenAI, check OPENAI_API_KEY. For Anthropic, set ANTHROPIC_API_KEY. For Ollama, ensure the Ollama server is running locally and set LLM_PROVIDER=ollama with the correct model name.
Frequently Asked Questions about Flexible GraphRAG
What is Flexible GraphRAG?
Flexible GraphRAG is a Model Context Protocol (MCP) server that python, llamaindex, langchain, docker compose: 15 property graph, 4 rdf , 10 vector, opensearch, elasticsearch, alfresco dbs. 13 data sources (9 auto-sync), kg auto-building, ontologies, llms, docling or llamaparse doc processing, graphrag, rag only, It connects AI assistants to external tools and data sources through a standardized interface.
How do I install Flexible GraphRAG?
Follow the installation instructions on the Flexible GraphRAG GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.
Which AI clients work with Flexible GraphRAG?
Flexible GraphRAG works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.
Is Flexible GraphRAG free to use?
Yes, Flexible GraphRAG is open source and available under the Apache-2.0 license. You can use it freely in both personal and commercial projects.
Flexible GraphRAG Alternatives — Similar Data Science & ML Servers
Looking for alternatives to Flexible GraphRAG? Here are other popular data science & ml servers you can use with Claude, Cursor, and VS Code.
Ultrarag
★ 5.6kA Low-Code MCP Framework for Building Complex and Innovative RAG Pipelines
RocketRide
★ 3.1k📇 🏠 - MCP server that exposes RocketRide AI pipelines as t
Aix Db
★ 2.1kAix-DB 基于 LangChain/LangGraph 框架，结合 MCP Skills 多智能体协作架构，实现自然语言到数据洞察的端到端转换。
NeMo Data Designer
★ 1.9k🎨 NeMo Data Designer: Generate high-quality synthetic data from scratch or from seed data.
PaperBanana
★ 1.7kOpen source implementation and extension of Google Research’s PaperBanana for automated academic figures, diagrams, and research visuals, expanded to new domains like slide generation.
MiniMax
★ 1.5kBridges MiniMax AI capabilities to the Model Context Protocol, enabling AI agents to perform image understanding, text-to-image generation, and speech synthesis. It provides a standardized interface for accessing MiniMax's core tools via JSON-RPC.
Browse More Data Science & ML MCP Servers
Explore all data science & ml servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.
Set Up Flexible GraphRAG in Your Editor
Choose your AI client for step-by-step setup instructions.
Quick Config Preview
Add this to your claude_desktop_config.json or .cursor/mcp.json
Ready to use Flexible GraphRAG?
Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.