LLM Graph Builder

Active

GitHub Jupyter Notebook Apache-2.0

Description

Extract entities and relations from unstructured text to build a Neo4j knowledge graph.

Key Features

Auto extraction — Entities and relations from PDF, web, documents
Multi-source — Wikipedia, local documents, S3 buckets
Multilingual — Chinese, English, Japanese, French, German, etc.
Visualization — Live graph structure display
Neo4j import — Push to Neo4j in one click
Multi-LLM — OpenAI, Anthropic, Gemini, and local models

Use Cases

💡 Build a graph index for enterprise knowledge bases.

💡 Turn scattered documents into a queryable knowledge network.

💡 Prepare data for GraphRAG.

Strengths & Limitations

✅ Strengths

• Actively maintained, recent updates
• High community interest (5.2k stars)
• Permissive open-source license (Apache-2.0)
• Established track record (2 years in production)

Quick Start

# Start docker-compose
docker compose up
# Open http://localhost:7474 in a browser
# Upload PDFs and auto-generate the graph

Visit GitHub

Related Projects

Neo4j GraphRAG

1.3k · Python

Active A

Official Neo4j GraphRAG Python SDK providing an integrated toolkit for knowledge graph construction, vector retrieval, and graph querying, supporting agent-driven graph retrieval-augmented generation workflows.

graphragneo4jknowledge-graph +2

· Official Neo4j GraphRAG Python SDK providing integrated toolkit for knowledge graph construction, vector retrieval, and graph querying
· Supports multiple LLM providers: OpenAI, Anthropic, Google, Cohere, Mistral, Bedrock, Ollama, and more
· SimpleKGPipeline for simplified knowledge graph construction supporting text and PDF input

View on GitHub

InstaGraph

3.6k · Python

Stale C

InstaGraph is a GPT-powered tool that turns text or URLs into color-coded knowledge graphs of entities and relationships, with optional Neo4j and FalkorDB persistence for explainable AI agent memory structures.

knowledge-graphllmneo4j +2

· Text to graph - extract entities and relationships from natural language or URLs
· Color-coded visualization - nodes and edges are color-coded for instant entity-type readability
· Responsive web UI - browser interface that works on desktop and mobile

View on GitHub

Unstract

7.2k · Python

Active A+

LLM-driven extraction of unstructured data, built for API deployments and ETL pipeline workflows. Automates document parsing, PDF extraction, and intelligent data processing with LLM-powered intelligence.

data-processingragpython +3

· Prompt Studio defines document extraction schemas with natural language
· Supports REST API deployment and ETL pipeline workflows
· MCP Server integration connecting to Claude and other AI agents

View full details → View on GitHub

RuoYi AI

5.7k · Java

Active A+

An out-of-the-box enterprise full-stack AI assistant platform with Langchain4j-based Supervisor multi-agent orchestration, RAG knowledge bases, visual workflow design, and multi-model management.

enterprisemulti-agentrag +2

· Multi-agent orchestration — Supervisor mode on Langchain4j with multiple decision models that combine tools and skills flexibly
· RAG knowledge management — local RAG with Milvus/Weaviate/Qdrant vector stores, document parsing, and retrieval QA
· Visual workflow designer — drag-and-drop node orchestration with SSE streaming for model calls, email, and human-approval nodes

View full details → View on GitHub

GraphRAGLightRAGRAG

GraphRAG vs LightRAG: Deep Comparison of Two Engineering Routes for Graph-Enhanced RAG

Systematic comparison of Microsoft GraphRAG (community detection + global summarization) and HKUDS LightRAG (dual-layer graph + incremental update) across index construction, retrieval paradigm, cost structure, query quality, and use cases. Includes a decision tree by data scale, query type, and ops budget.