Deploy and Host Langflow on Railway

Langflow is an open-source visual builder for AI applications. You connect LLMs, prompts, vector stores, tools and agents on a canvas, test them in the playground, and run them as a REST API, an MCP server for Claude, Cursor and other clients, or an embedded chat on your site. It works with OpenAI, Anthropic, Google, Ollama and many other providers.

About Hosting Langflow

Langflow is a Python web app that stores flows, users and credentials in a database and keeps uploaded files and caches on disk. In production it needs PostgreSQL, a persistent volume, a fixed secret key so saved credentials stay readable after restarts, and a login so strangers can't run code or spend your API credits.

This template deploys Langflow with PostgreSQL over Railway's private network, a volume for files, a generated secret key and admin password, and a pinned, tested version.

Common Use Cases

AI agents that use tools, APIs and your own data
RAG chatbots over documents, with vector stores such as Astra DB, Qdrant, Pinecone or pgvector
Multi-agent workflows and automations called from your app via API
Exposing flows as MCP tools for Claude, Cursor and other AI clients
Prototyping and testing prompts and models side by side

Dependencies for Langflow Hosting

PostgreSQL 18
Persistent volume for uploaded files and cache

Deployment Dependencies

Langflow: https://github.com/langflow-ai/langflow
Documentation: https://docs.langflow.org
Image: https://hub.docker.com/r/langflowai/langflow

Implementation Details

What you set: nothing. All secrets are generated.

After deploy:

Open the Langflow URL and log in as admin with the generated LANGFLOW_SUPERUSER_PASSWORD.
Add your provider keys under Settings → Global Variables.
Start from a template, then call your flow:

curl -X POST "https:///api/v1/run/" \
  -H "x-api-key: $LANGFLOW_API_KEY" -H "Content-Type: application/json" \
  -d '{"input_value": "Hello", "input_type": "chat", "output_type": "chat"}'

This template:

Requires a login (auto-login off) with a generated admin password, and turns off public sign-up
Lets only admins create custom components, which run Python on the server
Generates a persistent LANGFLOW_SECRET_KEY, so saved credentials survive restarts
Keeps flows in PostgreSQL on a volume, and uploaded files on a second volume
Connects to PostgreSQL over the private network
Pins Langflow 1.13.1 and PostgreSQL 18, logs at info level, and turns off telemetry
Has optional Langfuse tracing: set LANGFUSE_BASE_URL, LANGFUSE_PUBLIC_KEY and LANGFUSE_SECRET_KEY

Notes: don't change LANGFLOW_SECRET_KEY after the first deploy. Upgrade by changing the image tag; Langflow migrates the database on start. Langflow needs about 1–2 GB of RAM.

Why Deploy Langflow 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 Langflow 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.