Deploy and Host OpenCode Computer on Railway

Give OpenCode a persistent, private computer on Railway. This template deploys one OpenCode-ready Linux service with a persistent volume mounted at /data. Provider authentication, configuration, sessions, plugins, skills, and repositories survive redeploys. The template does not create a public domain or TCP proxy.

OpenCode Computer is an independent project. It is not built by, maintained by, or affiliated with the OpenCode team.

About Hosting OpenCode Computer

OpenCode Computer packages the pinned OpenCode CLI and server, Railway CLI, GitHub CLI, Git, tmux, ripgrep, Python, Node.js, and common development tools into a durable remote environment. User projects live in /data/workspace; OpenCode's complete XDG config, data, state, and cache hierarchy lives beneath /data/home; global skills are available at /data/skills.

The native OpenCode server is password-protected and listens only on 127.0.0.1:4096. Railway health uses a separate listener that exposes only /livez and /readyz. The future Agent Computer app will reach OpenCode through an authenticated bridge. Railway SSH remains available as an optional direct-access and recovery path.

Common Use Cases

• Keep OpenCode projects and sessions running away from a laptop.
• Continue work from desktop and mobile clients against one persistent environment.
• Use any OpenCode-supported model provider without rebuilding the image.
• Maintain durable agents, plugins, skills, configuration, and repositories.
• Use Railway SSH and the attached OpenCode TUI while integrating the Agent Computer UI.

Dependencies for OpenCode Computer Hosting

• A Railway account with capacity for one service and one persistent volume.
• An account or API credential for any provider supported by OpenCode.
• Optional: the Railway CLI and a local public SSH key for direct SSH access.

Deployment Dependencies

This template builds from the public m-abdelwahab/opencode-computer GitHub repository. Railway creates one service and attaches one volume at /data. No external database, public domain, or TCP proxy is required. The current OpenCode data store uses SQLite, so keep the service at one replica with the volume attached.

Why Deploy OpenCode Computer on Railway?

Railway supplies the persistent storage, compute, deployment lifecycle, metrics, and private access primitives needed by a remote coding computer. The template keeps those infrastructure details behind a single deployable unit while preserving Railway as an escape hatch.

For direct setup or troubleshooting, use Railway SSH:

railway ssh --project "$PROJECT_ID" --service opencode-computer

Then connect a provider and attach to the already-running private server:

opencode auth login
opencode-computer-attach

Provider-specific API keys can also be configured as Railway variables. Your private SSH keys remain on your local computer; Railway receives only public SSH keys when you opt into direct SSH access.