Deploy and Host DB-GPT on Railway

DB-GPT is an open source AI native data app framework. Ask questions about your databases in plain language and get SQL, charts and answers back, analyze Excel and CSV files, build knowledge bases for RAG, generate dashboards, and run multi agent data analysis and AWEL workflows. This template runs it against any OpenAI compatible API, with no GPU needed.

About Hosting DB-GPT

DB-GPT has no built in login: every visitor is treated as an admin, and the app can run SQL against connected databases and execute agent generated code. This template therefore never exposes it directly. The DB-GPT service runs on Railway's private network only, and a small Caddy gateway is the single public entry point, protected by a generated username and password and a cross site request guard. DB-GPT uses the official no GPU image eosphorosai/dbgpt-openai:v0.8.2, stores its metadata in SQLite and its vectors in embedded Chroma, both on one persistent volume, and keeps its database migrations on that same volume so redeploys and upgrades are safe.

Common Use Cases

Chat with your database: connect MySQL, PostgreSQL, ClickHouse, DuckDB or SQLite and ask questions in natural language (Text2SQL)
Spreadsheet analysis: upload Excel or CSV files and explore them conversationally
Knowledge base Q&A: index documents and URLs for retrieval augmented answers
Generative dashboards and data agents: let agents plan an analysis, write the SQL and render the charts

Dependencies for DB-GPT Hosting

An OpenAI compatible API key: required, used for both chat and embeddings. OpenAI is the default; OpenRouter, DeepSeek, a LiteLLM proxy and other compatible endpoints work by changing the base URL and model names
Your own databases: optional, added from the UI. An example Walmart sales database is preloaded

Deployment Dependencies

Implementation Details

Service	Image	Public	Volume
DB-GPT Gateway	`caddy:2.11-alpine`	yes	none
DB-GPT	`eosphorosai/dbgpt-openai:v0.8.2`	no	`/app/pilot/data`

First Steps After Deploy

Open the DB-GPT Gateway public URL
Log in with UI_USERNAME (default admin) and UI_PASSWORD, both on the gateway's Variables tab
Try Chat Data against the preloaded Walmart_Sales database
Add your own data sources and knowledge bases

Key Environment Variables

Variable	Service	Purpose
`OPENAI_API_KEY`	DB-GPT	Required LLM and embeddings key
`OPENAI_API_BASE`	DB-GPT	OpenAI compatible chat endpoint
`LLM_MODEL_NAME`	DB-GPT	Chat model, default `gpt-4o`
`EMBEDDING_MODEL_NAME` / `EMBEDDING_MODEL_API_URL`	DB-GPT	Embedding model and full endpoint URL
`DBGPT_ENCRYPT_KEY`	DB-GPT	Generated; encrypts stored secrets, never rotate
`UI_USERNAME` / `UI_PASSWORD`	DB-GPT Gateway	Login for the whole app

Notes

Single tenant: anyone with the gateway password has full admin access
The Hobby plan or higher is recommended; the Free plan's 0.5 GB of RAM is likely too small
Custom domain: attach it to the gateway and add it to DBGPT_CORS_ALLOWED_ORIGINS on the DB-GPT service

Why Deploy DB-GPT on Railway?

Railway is a singular platform to deploy your infrastructure stack. Railway will host your infrastructure so you don't have to deal with configuration, while allowing you to vertically and horizontally scale it.

By deploying DB-GPT on Railway, you are one step closer to supporting a complete full-stack application with minimal burden. Host your servers, databases, AI agents, and more on Railway.