How x3s Works
A plain-language overview of the x3s.host platform.
x3s.host is a managed hosting platform for AI agent runtimes. Here's what happens from sign-up to live agent.
The deployment flow
- You sign in and open the deploy page
- You choose a runtime, region, and billing plan
- x3s provisions a container on a dedicated server in your chosen region
- Your agent gets a unique public URL, credentials, and a browser sidecar
- Everything is accessible from your dashboard immediately
Deployment typically takes 30–60 seconds.
What's running behind your agent
Each agent is a small group of services running together:
- Runtime — the agent software you chose (OpenClaw, Moltis, etc.)
- Browser — a headless Chromium browser the runtime can use for web tasks
- Terminal — a shell session accessible directly from your dashboard
If Tailscale is enabled, a network sidecar is added and your agent gets a private IP on your Tailscale network.
Regions
Agents can be deployed in Europe (Helsinki) or the United States. Pick whichever is geographically closer to you or your data sources — it won't affect features, just latency.
What x3s manages for you
- Server provisioning and maintenance
- Container orchestration and restarts
- TLS certificates for your agent URL
- Browser sidecar lifecycle
- Monitoring and health checks
What you manage
- Your agent's secrets and API keys
- Runtime configuration and settings
- When to restart or redeploy
- Backups and snapshots before risky changes