Overview
What XO Cloud is, how it differs from XO Space, and the path from account to a running agent.
Cloud Space runs the open-source XO Space engine (cowork-api) on an isolated, remote cloud VM. You manage it directly from app.xo.builders—provisioning projects, attaching AI models, and accessing your workspace via browser IDE or API.
Local Space vs. Managed Cloud Space
Both options run the identical XO Space engine under the hood. Choose the deployment target that fits your workflow:
| Feature / Aspect | Self-Hosted Local Space | Managed Cloud Space |
|---|---|---|
| Core Engine | cowork-api (Local process) | cowork-api (Cloud VM container) |
| Primary Use Case | Local agent observability, offline work, zero cost | 1-click sandboxes, remote execution, team sharing |
| How You Install | `curl -fsSL quirq.ai/install | sh` or skill |
| IDE Access | Local Editor (VS Code, Cursor, Terminal) | Browser VS Code (code-server) & Desktop SSH/VS Code |
| Data & Telemetry | Stored 100% locally on your machine | Encrypted in isolated cloud project containers |
| Agent Support | OpenClaw, Claude Code, Hermes, Antigravity, etc. | OpenClaw, Claude Code, Hermes, Antigravity, etc. |
Engine Parity: Whether you run Space locally or on Cloud Space, every API endpoint, session log format, and connector functions identically.
Path on XO Cloud
Follow these guides in order for a first successful interaction.
1. Create an account
Sign up, choose a username, and select a plan.
2. Choose an agent
Pick OpenClaw, Claude Code, Hermes, XO Cowork, or Antigravity.
3. Launch your first agent
Create the XO project, wait for the workspace machine, complete a first test.
4. Operate the workspace
Monitor services, open the IDE, share the project.
Also in this section
After launch, Cloud docs continue with:
- Workspace configuration — Setup tab: models, data, channels
- AI Agent Guides — per-template Cloud quickstarts and deep dives