OpenComputer
OpenComputer is an open-source, local-first desktop agent that interprets the screen and operates applications with mouse and keyboard actions.
What OpenComputer is
OpenComputer is an open-source, local-first desktop agent that interprets the screen and operates applications with mouse and keyboard actions. It can click, drag, scroll, enter text, use shortcuts, open URLs, and wait for interface changes. It can use a local vision-language model through vLLM or a compatible hosted API; the endpoint must accept images and function tools and support the model’s tool-call format. Ask first is the default permission mode and requires approval for mouse, keyboard, and URL actions. Automatic mode skips individual prompts. The agent currently works only on the primary display. Linux X11 is best supported, while Wayland, Windows, and macOS are experimental; build targets include Linux AppImage and deb, Windows NSIS, and macOS DMG. The project calls itself early alpha and advises supervision, particularly for sensitive or irreversible actions. Screenshots, prompts, and tool results can remain on the user’s machine, but API keys are not backed by the operating-system keychain.
Good
- Runs locally and understands screen content.
- Offers approval prompts by default.
- Supports multiple mouse and keyboard actions.
- Screenshots and prompts can stay on-device.
- Available as Linux, Windows, and macOS build targets.
Watch out
- Project is in early alpha.
- Only operates on the primary display.
- Windows, macOS, and Wayland are experimental.
- API keys are not stored in the OS keychain.
Who it suits
It may suit users seeking local desktop automation who can supervise its actions and work within its current display and system limits. Users need a compatible model endpoint or a local model through vLLM.
Verdict
OpenComputer combines screen understanding with desktop controls and an approval-first default. Its early-alpha status, experimental platform support, and need for supervision are important considerations.
The facts
Each one read from OpenComputer's own pages, with the date it was checked.
What it does
- Action review
- Ask first is the default permission mode and requires approval for mouse, keyboard, and URL actions; Automatic mode runs actions without individual prompts.checked 7 Oct 2001
- API compatibility
- The configured endpoint must support image inputs, function tools, and the model’s tool-call format.checked 7 Oct 2001
- Credential storage
- API keys are stored in the per-user application-data directory with user-only file permissions, but are not yet backed by the operating-system keychain.checked 7 Oct 2001
- Desktop controls
- It supports clicking, dragging, scrolling, text entry, keyboard shortcuts, opening URLs, and waiting for UI changes.checked 7 Oct 2001
- Display limit
- The current agent operates only the primary display, and multi-monitor automation is on the roadmap.checked 7 Oct 2001
- Installer formats
- Build targets are Linux AppImage and deb, Windows NSIS, and macOS DMG.checked 7 Oct 2001
- Integrations
- The current model boundary uses an OpenAI-compatible Chat Completions API; dedicated Discord, search, and MCP integrations are listed as future work.checked 7 Oct 2001
- License
- OpenComputer is licensed under the MIT License, while third-party dependencies and model weights retain their own licenses.checked 7 Oct 2001
- Model compatibility
- The model endpoint must support image inputs, function tools, and the model's tool-call format.checked 7 Oct 2001
- Model options
- It can use a local vision-language model through vLLM or a compatible hosted API.checked 7 Oct 2001
- Privacy and security
- The README says screenshots, prompts, and tool results can stay on the user's machine, and that credentials, screenshots, model calls, and automation remain in Electron's main process.checked 7 Oct 2001
- Project maturity
- The project describes itself as early alpha and warns that desktop actions should be supervised, especially around sensitive or irreversible workflows.checked 7 Oct 2001
- Purpose
- OpenComputer is an open-source, local-first desktop agent that understands the screen and operates applications through mouse and keyboard interactions.checked 7 Oct 2001
- Security design
- The Electron renderer is sandboxed with Node.js integration disabled, while privileged model and desktop automation behavior runs in the main process.checked 7 Oct 2001
- Security limitations
- The project says its controls are defense in depth rather than a complete sandbox, and Automatic mode lets the model act with the user's desktop privileges.checked 7 Oct 2001
- Security reporting
- The security policy asks users to report vulnerabilities privately and says reports should receive an initial response within seven days, though response times may vary because it is a volunteer project.checked 7 Oct 2001
- Supported systems
- Build targets are Linux AppImage/deb, Windows NSIS, and macOS DMG; Linux X11 is best supported, while Wayland, Windows, and macOS are experimental.checked 7 Oct 2001
- What it does
- OpenComputer is an open-source desktop agent that understands the screen and can operate applications with mouse and keyboard interactions.checked 7 Oct 2001
Where it runs
- Approvals
- Ask first is the default permission mode and requires approval for mouse, keyboard, and URL actions.checked 7 Oct 2001
The company
- Maker
- The linked GitHub profile identifies the maker as Andrew Krikorian and lists California.checked 7 Oct 2001
Alternatives to OpenComputer
- 1
Claude Code
Free planAndroidiOSWebWindows7.4From $17 a month - 2
Gemini
Free planAndroidiOSWebWindows6.9Free plan - 3
Computer Use OOTB
Free planWebmacOSSelf-hostedWindows6.8Free plan - 4
Lapu AI
Free planmacOSWindows6.7From €16 a month - 5
QuickDesk
Free planAndroidWebAPImacOS6.5Free plan - 6
Contop
AndroidiOSLinuxmacOS6.4