---
title: "Deploy LobeChat | Multi-Model AI/LLM Support"
description: "Self Host Lobechat. Private AI chat with 40+ providers, RAG, plugins & more"
category: "AI/ML"
url: https://railway.com/deploy/lobechat
---

# Deploy LobeChat | Multi-Model AI/LLM Support

Self Host Lobechat. Private AI chat with 40+ providers, RAG, plugins & more

**[Deploy LobeChat | Multi-Model AI/LLM Support on Railway](https://railway.com/template/lobechat)**

- **Creator:** Heimdall
- **Category:** AI/ML
- **Total deploys:** 43

## Template content

### PostgreSQL https://cdn.jsdelivr.net/gh/homarr-labs/dashboard-icons/svg/postgresql.svg

- **Image:** paradedb/paradedb:latest-pg17

### RustFS Init https://cdn.jsdelivr.net/gh/homarr-labs/dashboard-icons/svg/rustfs.svg

- **Image:** alpine:latest
- **Start command:** `/bin/sh -c 'wget -q https://dl.min.io/client/mc/release/linux-amd64/mc -O /tmp/mc && chmod +x /tmp/mc && for i in 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20; do /tmp/mc alias set rustfs http://RustFS.railway.internal:9000 $RUSTFS_ACCESS_KEY $RUSTFS_SECRET_KEY 2>/dev/null && break; sleep 5; done && /tmp/mc mb rustfs/lobe --ignore-existing && printf "{\"Statement\":[{\"Effect\":\"Allow\",\"Principal\":{\"AWS\":[\"*\"]},\"Action\":[\"s3:GetObject\"],\"Resource\":[\"arn:aws:s3:::lobe/*\"]}],\"Version\":\"2012-10-17\"}" > /tmp/bp.json && /tmp/mc anonymous set-json /tmp/bp.json rustfs/lobe && echo Init done && sleep infinity'`

### RustFS https://cdn.jsdelivr.net/gh/homarr-labs/dashboard-icons/svg/rustfs.svg

- **Image:** rustfs/rustfs:latest
- **Start command:** `/bin/sh -c 'mkdir -p /data/lobe && exec rustfs --access-key $RUSTFS_ACCESS_KEY --secret-key $RUSTFS_SECRET_KEY /data'`
- **Public domain:** Yes

### LobeChat https://cdn.jsdelivr.net/gh/homarr-labs/dashboard-icons/svg/lobe-chat.svg

- **Image:** lobehub/lobehub
- **Public domain:** Yes

### Redis https://cdn.sanity.io/images/sy1jschh/production/0ce0bfdcfbdbf69662b1116671f97c2dd788b655-157x157.svg

- **Image:** redis:8.2.1
- **Start command:** `/bin/sh -c "rm -rf $RAILWAY_VOLUME_MOUNT_PATH/lost+found/ && exec docker-entrypoint.sh redis-server --requirepass $REDIS_PASSWORD --save 60 1 --dir $RAILWAY_VOLUME_MOUNT_PATH"`

## Documentation

![LobeChat logo](https://picx.zhimg.com/v2-709d52943998be9d1d9edc45f8d3ea11_720w.jpg?source=172ae18b)

# Deploy and Host LobeChat on Railway

[![Deploy on Railway](https://railway.com/button.svg)](https://railway.com/deploy/lobechat?referralCode=QXdhdr&utm_medium=integration&utm_source=template&utm_campaign=generic)

Deploy LobeChat on Railway to run your own private AI chat platform with multi-model support, knowledge bases, and plugin extensibility. Self-host LobeChat with full control over your data and AI provider keys — no vendor lock-in.

This Railway template deploys a production-ready LobeChat stack with PostgreSQL (via ParadeDB with pgvector), Redis for caching and sessions, and RustFS for S3-compatible file storage. All services are pre-configured with secure inter-service networking and persistent volumes.

## Getting Started with LobeChat on Railway

After Railway finishes deploying all five services, open your LobeChat URL to reach the sign-in page. LobeChat 2.0 uses Better Auth — register with an email and password to create your first account. Once logged in, navigate to **Settings > Model Provider** and add your API key for at least one provider (OpenAI, Anthropic Claude, Google Gemini, or any of the 40+ supported providers). Send your first message to verify the connection works.

To enable file uploads and the knowledge base, your RustFS storage is already configured. Upload a document in any chat to test RAG capabilities. Explore the Plugin Store and Agent Marketplace to extend functionality beyond basic chat.

## About Hosting LobeChat

LobeChat is an open-source AI chat framework with 70,000+ GitHub stars, built with Next.js. It serves as a private alternative to ChatGPT with full multi-model support.

- **Multi-Provider AI**: Connect 40+ providers — OpenAI, Claude, Gemini, DeepSeek, Ollama, AWS Bedrock, Azure, Mistral, Groq, and more
- **Knowledge Base (RAG)**: Upload documents, build searchable knowledge bases with pgvector-powered retrieval
- **Plugin Ecosystem**: 10,000+ skills via Function Calling, MCP marketplace integration
- **Multi-Modal**: Vision (image recognition), text-to-speech, speech-to-text, image generation
- **Better Auth**: Built-in authentication with email/password, OAuth, SSO support
- **Artifacts & Thinking**: Advanced reasoning visualization and interactive code artifacts

## Why Deploy LobeChat on Railway

- One-click deployment with PostgreSQL, Redis, and S3 storage pre-configured
- Private networking between services — database never exposed to the internet
- Persistent volumes for database and file storage survive redeploys
- Scale services independently — add more RAM to LobeChat without touching the database
- No Docker knowledge required — Railway handles containers, networking, and TLS

## Common Use Cases for Self-Hosted LobeChat

- **Team AI Assistant**: Deploy a shared ChatGPT alternative with centralized API key management and user accounts
- **RAG-Powered Knowledge Base**: Upload internal docs and build a searchable AI assistant for your organization
- **Multi-Model Testing**: Compare responses from GPT-4, Claude, Gemini, and open-source models side-by-side
- **Custom Agent Development**: Build specialized AI agents with plugins, tools, and MCP integrations for domain-specific workflows

## Dependencies for LobeChat on Railway

- **LobeChat** — `lobehub/lobehub` — Main application (Next.js 16, port 8080)
- **PostgreSQL** — `paradedb/paradedb:latest-pg17` — Database with pgvector extension for RAG
- **Redis** — Railway managed — Session storage, caching, rate limiting
- **RustFS** — `rustfs/rustfs:latest` — S3-compatible object storage for file uploads
- **RustFS Init** — `alpine:latest` — One-shot bucket initialization using `minio/mc`

### Environment Variables Reference for LobeChat

| Variable | Service | Description |
|----------|---------|-------------|
| `DATABASE_URL` | LobeChat | PostgreSQL connection string |
| `REDIS_URL` | LobeChat | Redis connection for sessions/cache |
| `S3_ENDPOINT` | LobeChat | RustFS public URL for file storage |
| `S3_BUCKET` | LobeChat | S3 bucket name (`lobe`) |
| `KEY_VAULTS_SECRET` | LobeChat | Encryption key for user API key vaults |
| `AUTH_SECRET` | LobeChat | Session token signing secret |
| `APP_URL` | LobeChat | Public-facing application URL |
| `POSTGRES_DB` | PostgreSQL | Database name |
| `POSTGRES_PASSWORD` | PostgreSQL | Database password |
| `PGDATA` | PostgreSQL | Data directory (subdirectory of volume) |
| `RUSTFS_ACCESS_KEY` | RustFS | S3 access key |
| `RUSTFS_SECRET_KEY` | RustFS | S3 secret key |

### Deployment Dependencies

- **Runtime**: Node.js 22 (bundled in Docker image)
- **Docker Hub**: [lobehub/lobehub](https://hub.docker.com/r/lobehub/lobehub)
- **GitHub**: [lobehub/lobe-chat](https://github.com/lobehub/lobe-chat) (70k+ stars)
- **Docs**: [lobehub.com/docs/self-hosting](https://lobehub.com/docs/self-hosting)

## Hardware Requirements for Self-Hosting LobeChat

| Resource | Minimum | Recommended |
|----------|---------|-------------|
| CPU | 1 vCPU | 2 vCPU |
| RAM | 1 GB (app) + 512 MB (PostgreSQL) | 2 GB (app) + 1 GB (PostgreSQL) |
| Storage | 2 GB (app) + 5 GB (database) | 10 GB (database) + 20 GB (files) |
| Runtime | Docker 20+ or Node.js 20+ | Docker Compose recommended |

RAM scales with concurrent users and knowledge base size. PostgreSQL with pgvector benefits from additional memory for vector index caching.

## Self-Hosting LobeChat with Docker Compose

The fastest way to self-host LobeChat is with Docker Compose. Clone the official repository and use the deploy configuration:

```
git clone https://github.com/lobehub/lobe-chat.git
cd lobe-chat/docker-compose/deploy
cp .env.example .env
# Edit .env — set KEY_VAULTS_SECRET, AUTH_SECRET, POSTGRES_PASSWORD
bash setup.sh  # generates secrets automatically
docker compose up -d
```

For a minimal single-service setup without persistent storage:

```
docker run -d \
  -p 3210:8080 \
  -e KEY_VAULTS_SECRET="$(openssl rand -base64 32)" \
  -e AUTH_SECRET="$(openssl rand -base64 32)" \
  -e DATABASE_URL="postgresql://user:pass@host:5432/lobechat" \
  lobehub/lobehub
```

## How Much Does LobeChat Cost to Self-Host?

LobeChat is free and open-source under the Apache 2.0 license. There are no paid tiers for the self-hosted version — you get the full feature set. The only costs are infrastructure (server, database, storage) and AI provider API keys. On Railway, expect approximately $5–15/month for the full stack depending on usage. LobeHub also offers a managed cloud service with credit-based pricing for teams that prefer not to self-host.

## LobeChat vs Open WebUI vs LibreChat

| Feature | LobeChat | Open WebUI | LibreChat |
|---------|----------|------------|-----------|
| License | Apache 2.0 | MIT | MIT |
| AI Providers | 40+ (API-based) | Ollama + OpenAI | 20+ |
| Knowledge Base (RAG) | Built-in with pgvector | Built-in | Plugin-based |
| Plugin System | 10,000+ skills | Pipelines | Plugins |
| Auth | Better Auth (SSO, OAuth) | Built-in | OAuth |
| Local LLMs | Via Ollama integration | Native Ollama | Via Ollama |
| GitHub Stars | 70k+ | 80k+ | 20k+ |

LobeChat excels at multi-provider API-based workflows with the richest plugin ecosystem. Open WebUI is the stronger choice for local LLM-first deployments with Ollama.

## FAQ

**What is LobeChat and why self-host it?**
LobeChat is an open-source AI chat framework that lets you build a private ChatGPT-like interface. Self-hosting gives you full control over data privacy, API key management, and the ability to connect any AI provider without third-party intermediaries.

**What does this Railway template deploy for LobeChat?**
This template deploys five services: the LobeChat application (lobehub/lobehub), PostgreSQL with pgvector for database and vector search, Railway-managed Redis for caching, RustFS for S3-compatible file storage, and a one-shot init service that creates the storage bucket.

**Why does LobeChat need PostgreSQL with pgvector on Railway?**
PostgreSQL stores user accounts, chat history, settings, and plugin configurations. The pgvector extension enables the Knowledge Base feature — when you upload documents, LobeChat creates vector embeddings and stores them in PostgreSQL for semantic search (RAG).

**How do I add my OpenAI or Claude API key to self-hosted LobeChat?**
After logging in, go to **Settings > Model Provider** and enter your API key for each provider. LobeChat encrypts stored keys using the `KEY_VAULTS_SECRET` environment variable. You can add keys for OpenAI, Anthropic Claude, Google Gemini, and 40+ other providers simultaneously.

**Can I use LobeChat with local LLMs like Ollama on Railway?**
LobeChat supports Ollama as a model provider, but running Ollama requires GPU access which Railway does not currently provide. You can connect LobeChat on Railway to an Ollama instance running on a separate GPU server by setting the Ollama API endpoint in the model provider settings.

**How do I enable file uploads and knowledge base in self-hosted LobeChat?**
This template pre-configures RustFS as S3-compatible storage. File uploads and knowledge base features work out of the box — upload documents in any chat conversation or create dedicated knowledge bases from the sidebar.


## 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/lobechat