Deploy AI Coding Workstation: Claude Code, Codex, Gemini
Browser VS Code and terminal with Claude Code, Codex and Gemini CLI
AI Workstation
Just deployed
/data
Deploy and Host AI Coding Workstation: Claude Code, Codex, Gemini on Railway
A private cloud workstation for AI coding agents that you open in any browser, including on your phone. You get VS Code in the browser, a lightweight web terminal, and Claude Code, OpenAI Codex, Gemini CLI and OpenCode pre-installed, all behind one password and backed by a persistent volume.
About Hosting AI Coding Workstation: Claude Code, Codex, Gemini
The template builds one Ubuntu 24.04 service from a public Dockerfile. Caddy serves everything on a single HTTPS port: VS Code (code-server) at / with its own password page, and a ttyd web terminal at /terminal with basic auth (user dev). A random PASSWORD is generated for you; copy it from the service Variables. A Railway volume is mounted at /data, so your home folder, agent logins, shell history and projects survive redeploys. The agents update themselves in the background on every boot, and the process supervisor restarts the container if anything exits.
Common Use Cases
- Run Claude Code or Codex 24/7 inside tmux while your laptop is closed, and check in from your phone
- A full coding environment on an iPad, Chromebook or borrowed computer, with nothing to install
- A disposable sandbox where agents can install packages and run code away from your own machine
Dependencies for AI Coding Workstation: Claude Code, Codex, Gemini Hosting
- No other services: one container plus a volume
- A Claude, ChatGPT or Google account to sign in to the agents, or API keys in the optional variables
Implementation Details
Sign in once from the terminal and the login is saved on the volume:
claude # opens a login link, or set ANTHROPIC_API_KEY
codex login --device-auth # or set OPENAI_API_KEY
gemini # "Login with Google", or set GEMINI_API_KEY
Projects live in /data/projects. Tools included: Node.js 22, Python 3 with uv and pipx, git, GitHub CLI, tmux, ripgrep, fd, jq and build-essential. Set GITHUB_TOKEN to push to your repositories. Give the service 2 GB of memory or more when running several agents at once. Source: https://github.com/nerminsabanovic31/railway-ai-coding-workstation
Why Deploy AI Coding Workstation: Claude Code, Codex, Gemini 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 AI Coding Workstation: Claude Code, Codex, Gemini 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
AI Workstation
nerminsabanovic31/railway-ai-coding-workstation
