Documentation

Concepts and a practical path from first credentials to a controllable machine.

Introduction

A Cloud Computer AI computer is a full graphical environment. Your software creates a machine, receives stream or control endpoints, and drives the session through screenshots, pointer events, keystrokes, and optional shell commands. Cloud Computer AI is deliberately not the agent itself; it is the computer on which the agent works.

Quickstart

  1. Create an account and open the dashboard.
  2. Launch a computer from Computers or via the Cloud Computer AI API.
  3. Attach your agent loop: capture the display, pass it to a model, apply actions.
  4. Open the live stream when human observation is useful.
  5. Stop or delete the machine when work is complete.

Core API ideas

Typical flows include creating a computer, polling status, capturing the display, sending pointer and keyboard events, running shell commands, and deleting the instance when finished. The Express console at the API host also exposes workspace, templates, secrets, and E3B monitoring.

E3B monitoring

E3B is Cloud Computer AI’s fleet-monitoring surface: concurrent sandboxes, resource overview, events, and usage estimates. Use it from the dashboard sidebar or the classic console when you need operational visibility across machines.

Cloud Computer AI
© 2026 Cloud Computer AI. Cloud computers for intelligent agents · cloudcomputerai.com
Cloud Computer AI