Deploy MindsDB
MindsDB AI SQL engine: pinned v26.1.0, login on, data kept on a volume
Just deployed
/root/mdb_storage
Deploy and Host MindsDB on Railway
MindsDB is an open-source AI query engine. You connect databases, warehouses, SaaS apps and files once, then query them together in SQL, build knowledge bases for semantic search, and run agents and scheduled jobs over live data. It serves a web Studio, an HTTP API, an MCP server and a MySQL-compatible API.
About Hosting MindsDB
This template deploys MindsDB v26.1.0 from the official mindsdb/mindsdb image, pinned by version and digest, as one service built from a public GitHub repository. Login is always on: the password, the API token and the session key are generated for each deploy, so the Studio, API, MCP and MySQL endpoints are never open. Everything MindsDB stores (its catalogue of projects, connections, models, views, jobs, agents and knowledge bases, plus uploaded files) lives on a Railway volume and survives redeploys. The Studio ships inside the image, so the first boot downloads nothing. A health check on /api/util/ping gates each deploy, and a TCP proxy publishes the MySQL API for BI tools.
Common Use Cases
- Query Postgres, MySQL, Snowflake, Salesforce, HubSpot or files together in one SQL statement, without an ETL pipeline
- Give Claude, Cursor or another MCP client grounded, read-only access to your company data
- Build knowledge bases and agents that answer plain-English questions over documents and tables
Dependencies for MindsDB Hosting
- A Railway volume for MindsDB's data (included)
- An API key for an LLM provider such as OpenAI, only if you use agents or knowledge bases
Deployment Dependencies
- MindsDB release v26.1.0: https://github.com/mindsdb/mindsdb/releases/tag/v26.1.0
- MindsDB Docker docs: https://docs.mindsdb.com/setup/self-hosted/docker
- MindsDB environment variables: https://docs.mindsdb.com/setup/environment-vars
- Template source: https://github.com/aalfath/mindsdb-railway-template
Implementation Details
After the deploy is green:
- Open the service's public URL and sign in with
MINDSDB_USERNAME(defaultmindsdb) andMINDSDB_PASSWORD, both in the service's Variables tab. - Connect a data source in the Studio's SQL editor, for example
CREATE DATABASE my_pg WITH ENGINE = 'postgres', PARAMETERS = {...};, then query it withSELECT. - For AI clients, use the MCP endpoint
https://YOUR-DOMAIN/mcp/streamable(or/mcp/sse) with the headerAuthorization: Bearer MINDSDB_AUTH_TOKEN(the variable's value). For the HTTP API, send the same header tohttps://YOUR-DOMAIN/api/sql/query. - For BI tools and MySQL clients, use the TCP proxy host and port shown under the service's Settings → Networking, with the same user and password.
Optional: set MINDSDB_DEFAULT_LLM_API_KEY for agents and knowledge bases, MINDSDB_DB_CON to keep the catalogue in a Postgres database, or KB_PGVECTOR_URL to store knowledge bases in Postgres with pgvector.
Updates: the service builds from the template's GitHub repository, where the image is pinned in the Dockerfile. When a new MindsDB version is merged there, Railway shows you an update notice and you choose when to apply it. MindsDB migrates its data on start, and the changes are listed in the repository's CHANGELOG.md.
Why Deploy MindsDB 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 MindsDB 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.
Template Content

