---
title: "Deploy lightrag"
description: "One-click Graph RAG • Knowledge Graph + Vector Search • Web UI included"
category: "AI/ML"
url: https://railway.com/deploy/lightrag
---

# Deploy lightrag

One-click Graph RAG • Knowledge Graph + Vector Search • Web UI included

**[Deploy lightrag on Railway](https://railway.com/template/lightrag)**

- **Creator:** Rohan Sharma's Projects
- **Category:** AI/ML
- **Total deploys:** 38

## Template content

### lightrag-railway

- **Source:** https://github.com/rohansx/lightrag-railway

## Documentation

# Deploy and Host LightRAG on Railway

One-click deployment of LightRAG - the most popular Graph RAG framework with 25K+ GitHub stars. Featured in EMNLP 2025.

## About Hosting LightRAG

LightRAG is a graph-based Retrieval-Augmented Generation system that outperforms traditional RAG by building knowledge graphs from your documents. Unlike simple vector search, LightRAG understands relationships between entities, delivering smarter and more contextual answers.

This template includes the full LightRAG server with a built-in Web UI for document management, chat interface, and interactive knowledge graph visualization.

## Why Deploy LightRAG on Railway

- **One-click setup**: No Docker configuration or server management required
- **Built-in Web UI**: Upload documents, chat, and visualize your knowledge graph instantly
- **Smarter than traditional RAG**: Graph-based retrieval captures entity relationships, not just text similarity
- **Multiple LLM providers**: Works with OpenAI, Groq, OpenRouter, Azure, and AWS Bedrock
- **Production ready**: Health checks, automatic restarts, and scalable infrastructure
- **Cost effective**: Pay only for what you use with Railway's usage-based pricing

## Common Use Cases

- **Research & Academia**: Analyze research papers and discover connections across literature
- **Documentation Q&A**: Build intelligent search for technical documentation
- **Legal Document Review**: Extract entities and relationships from contracts and legal texts
- **Customer Support**: Create knowledge bases that understand context and relationships
- **Business Intelligence**: Query company documents with natural language
- **Personal Knowledge Management**: Organize and query your notes, books, and articles

## Dependencies for LightRAG

LightRAG requires an LLM provider for text generation and embeddings. The template supports multiple providers out of the box.

### Deployment Dependencies

- **LLM Provider** (Required): OpenAI API key recommended. Also supports Groq, OpenRouter, Azure OpenAI, and AWS Bedrock
- **Embedding Model** (Required): Uses OpenAI text-embedding-3-small by default (1536 dimensions)
- **Storage**: Railway provides ephemeral storage. Add a Railway Volume mounted to `/app/data` for persistent storage
- **Memory**: Minimum 512MB RAM recommended, 1GB+ for larger document collections

### Environment Variables

| Variable | Required | Default | Description |
|----------|----------|---------|-------------|
| LLM_BINDING_API_KEY | Yes | - | Your LLM provider API key |
| LLM_MODEL | No | gpt-4o-mini | Model for text generation |
| EMBEDDING_MODEL | No | text-embedding-3-small | Model for embeddings |
| EMBEDDING_DIM | No | 1536 | Embedding dimensions |

### Quick Start

1. Deploy this template
2. Set your `LLM_BINDING_API_KEY` environment variable
3. Generate a public domain in Settings → Networking
4. Access the Web UI and upload your first document

### Links

- [LightRAG GitHub](https://github.com/HKUDS/LightRAG) - 25K+ stars
- [Research Paper](https://arxiv.org/abs/2410.05779) - EMNLP 2025

## Similar templates

- [Chat Chat](https://railway.com/deploy/-WWW5r) — Chat Chat, your own unified chat and search to AI platform.
- [stella](https://railway.com/deploy/stella) — Self-host stella with web, API, Postgres, Redis, and object storage.
- [Hermes Agent | OpenClaw Alternative with Dashboard](https://railway.com/deploy/hermes-agent-or-openclaw-alternative-wit) — Self-Hosted Hermes AI Agent for Telegram, Discord & Slack

Open this page in a browser: https://railway.com/deploy/lightrag