Deploy and Host Sim Studio on Railway

Sim Studio is an open-source platform for building and running AI agent workflows. You connect LLMs, tools, APIs and knowledge bases on a visual canvas, then deploy the workflows as APIs, chat apps, webhooks, schedules or MCP tools. It supports OpenAI, Anthropic, Gemini and many other model providers, plus over 100 integrations.

About Hosting Sim Studio

Self-hosting Sim Studio involves several cooperating services. These are the Next.js app, a realtime WebSocket server for live collaboration, a one-off migrations job, and a cron scheduler for background jobs, schedules and polling triggers. Data is stored in PostgreSQL with pgvector for knowledge-base embeddings, Redis handles pub/sub and caching, and S3-compatible storage holds uploaded files. This template wires everything together over Railway's private network. It generates unique secrets for every install, sets up healthchecks and restart policies, and connects a Railway bucket for file storage. Only the app and the realtime server are public. Postgres, Redis and the cron service stay private.

Common Use Cases

Building AI agents and multi-step LLM workflows with a visual editor
Automating business processes with scheduled, webhook or email-triggered workflows across tools like Gmail, Slack, HubSpot and Google Workspace
Running private RAG and knowledge-base assistants on your own data and infrastructure
Exposing your workflows to Claude, Cursor and other AI clients through the built-in MCP server

Dependencies for Sim Studio Hosting

PostgreSQL 17 with pgvector (pgvector/pgvector:pg17)
Redis
S3-compatible object storage (Railway Bucket)
Sim Studio images: simstudio, realtime, migrations and cron from ghcr.io/simstudioai

Implementation Details

Services included: simstudio (app, port 3000), realtime (WebSocket, port 3002), migrations (runs once, then exits), cron (background jobs), Postgres (pgvector), Redis and a Railway Bucket. All images track latest.

After deploying:

Open the simstudio domain and sign up. The first account becomes yours. Signup is open by default.
Add your model provider key as a workspace secret (Settings > Secrets), for example OPENAI_API_KEY, and reference it in Agent blocks as {{OPENAI_API_KEY}}. On self-hosted Sim, Agent blocks don't use a server-level provider key.

Optional variables (set on the simstudio service):

Variable	Purpose
`ALLOWED_LOGIN_EMAILS`	Comma-separated emails allowed to sign up. Recommended, to lock down your instance.
`ALLOWED_LOGIN_DOMAINS`	Allow signup for whole email domains, e.g. `company.com`
`DISABLE_REGISTRATION`	Set `true` to block all new signups after your account exists
`OPENAI_API_KEY`	Enables knowledge-base document embeddings
`COPILOT_API_KEY`	Enables the built-in Chat/Copilot (key from sim.ai). Also set `NEXT_PUBLIC_CHAT_DISABLED=false`.

MCP server: available at https:///api/mcp, with OAuth sign-in or a Sim API key.

Notes:

Chat is disabled by default because it needs a Sim Copilot key. An OpenAI key can't be used for Chat.
Railway Buckets require S3_FORCE_PATH_STYLE=true and AWS_REGION=auto for browser uploads. Both are preconfigured.
Email (invites, password reset, verification) and OAuth sign-in providers aren't configured by default.
PII detection (Guardrails PII block) is an optional add-on. Add a service from ghcr.io/simstudioai/pii:latest (port 5001, private) and set PII_URL=http://${{pii.RAILWAY_PRIVATE_DOMAIN}}:5001 and INTERNAL_API_BASE_URL=http://${{RAILWAY_PRIVATE_DOMAIN}}:3000 on simstudio. It needs about 4 GB of RAM or more.

Why Deploy Sim Studio 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 Sim Studio 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.