# Build on the AI-Ready Knowledge Graph

Everything you need to connect, query, and build applications on Fluree — from quickstart guides to API reference to open-source tools.

## Get Started

### Quickstart

Zero to first query in 5 minutes. Connect a data source, run a query, see results.

### Fluree Solo

Free serverless knowledge graph. No infrastructure. Full Intelligence Stack included.

### MCP Server

Connect any LLM to your knowledge graph via Model Context Protocol. Claude, ChatGPT, any MCP client.

## Documentation

### Organized by what you’re trying to accomplish

Not by product name.

### Connect Your Data
- Connector setup guides for top 20 sources
- PostgreSQL, Salesforce, S3, Snowflake
- File upload (CSV, JSON, PDF, Parquet)

### Query Your Graph
- SPARQL reference
- FlureeQL syntax
- REST API reference
- Natural language query setup

### Build AI Applications
- MCP server configuration
- LangChain integration
- LlamaIndex connector
- Agent framework patterns

### Model Your Data
- Ontology creation guide
- ITM taxonomy management
- Schema design patterns
- JSON-LD / RDF primer

### Secure Your Data
- Access policy configuration
- Role-based access control
- Audit trail & compliance
- Data-centric security guide

### Deploy & Operate
- Docker deployment
- Cloud deployment (AWS, Azure, GCP)
- Monitoring & scaling
- Backup & recovery

## Downloads

### Fluree DB v3.0

The core knowledge graph database engine.

- Linux (x64) `fluree-3.0-linux-x64.tar.gz`
- macOS (ARM) `fluree-3.0-macos-arm64.tar.gz`
- macOS (x64) `fluree-3.0-macos-x64.tar.gz`
- Docker `docker pull fluree/server:3.0`

Quick start
`docker run -p 8090:8090 fluree/server:3.0`

### MCP Server

Connect any MCP-compatible AI client to your Fluree knowledge graph.

`fluree memory mcp-install`

Works with Claude Desktop, ChatGPT, Amazon Bedrock, Cursor, and any MCP-compliant framework.

### SDKs & Libraries
- JavaScript / TypeScript `npm install @fluree/sdk`
- Python `pip install fluree-sdk`
- Java `Maven: com.fluree/fluree-sdk`
- LangChain `pip install fluree-langchain`
- LlamaIndex `pip install fluree-llamaindex`

Previous versions: github.com/fluree/db/releases · License: AGPL v3 (open source) or commercial license.

## Fluree Labs — Build in Your Browser

### Ontology Builder

Visually design and export JSON-LD ontologies for your knowledge graph. Drag-and-drop classes, properties, and relationships. Export as JSON-LD, Turtle, or OWL.

### Knowledge Graph Builder

Build and explore knowledge graphs interactively. Add entities, create relationships, run queries, and visualize your graph — all in the browser.

### Chat with Our Codebase

Ask questions about Fluree’s source code and get AI-powered answers grounded in the actual codebase. Understand architecture decisions and explore the code without cloning.

## Benchmarks

### LLM Accuracy Benchmark

Comparing GraphRAG vs traditional vector RAG — measuring answer accuracy, hallucination rates, and provenance quality across LLM-powered Q&A.

4× improvement in zero-shot accuracy. 95%+ with semantic enrichment.

### RDF Performance Benchmark

Performance benchmarks for Fluree’s SPARQL query engine — insert throughput, query latency, and scalability across standard RDF workloads.

## Community

### Discord

Join 800+ developers building with Fluree. Real-time help from the engineering team.

### GitHub

Source code, issues, pull requests, and example repositories.

### Fluree Labs Blog

Technical deep-dives from the engineering team. FlureeQL vs SPARQL, immutable architectures, semantic RAG.

## API Reference

### REST API Preview

- POST `/fluree/query` Execute a SPARQL or FlureeQL query
- POST `/fluree/transact` Write data to the knowledge graph
- GET `/fluree/dataset` List available datasets
- GET `/fluree/health` Health check

## Start Building.

Fluree Solo is free. The docs are comprehensive. The Discord is active. The MCP server takes 30 seconds to configure. What are you building?