Skip to content
SPECIFICTION
Get listed

OpenComputer

github.com

WindowsmacOSLinuxAPI
Visit github.com
Spec score 7.2 #18 of 24 in AI Computer Use Agents

What it is

OpenComputer is a free, MIT-licensed open-source desktop agent that runs locally and uses screen understanding to operate applications through mouse and keyboard actions. It can click, drag, scroll, enter text, use shortcuts, open URLs, and wait for interface changes. The agent can connect to a local vision-language model through vLLM or to a hosted API; the endpoint must accept image inputs, function tools, and the model’s tool-call format. Ask first is the default permission mode and requests approval for mouse, keyboard, and URL actions, while Automatic mode runs them without individual prompts. Build targets include Linux AppImage and deb packages, Windows NSIS, and macOS DMG. Linux X11 is best supported; Wayland, Windows, and macOS are experimental. The current agent operates only on the primary display. Screenshots, prompts, and tool results can remain local. API keys use user-only file permissions but are not stored in the operating-system keychain. The project describes itself as early alpha and recommends supervision for sensitive workflows.

Spec sheet

What it does
OpenComputer is an open-source desktop agent that understands the screen and can operate applications with mouse and keyboard interactions.Read 7 Oct 2026 · github.com
Purpose
OpenComputer is an open-source, local-first desktop agent that understands the screen and operates applications through mouse and keyboard interactions.Read 7 Oct 2026 · github.com
Integrations
The current model boundary uses an OpenAI-compatible Chat Completions API; dedicated Discord, search, and MCP integrations are listed as future work.Read 7 Oct 2026 · github.com
License
OpenComputer is licensed under the MIT License, while third-party dependencies and model weights retain their own licenses.Read 7 Oct 2026 · github.com
Maker
The linked GitHub profile identifies the maker as Andrew Krikorian and lists California.Read 7 Oct 2026 · github.com
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.Read 7 Oct 2026 · github.com
API compatibility
The configured endpoint must support image inputs, function tools, and the model’s tool-call format.Read 7 Oct 2026 · github.com
Approvals
Ask first is the default permission mode and requires approval for mouse, keyboard, and URL actions.Read 7 Oct 2026 · github.com
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.Read 7 Oct 2026 · github.com
Desktop controls
It supports clicking, dragging, scrolling, text entry, keyboard shortcuts, opening URLs, and waiting for UI changes.Read 7 Oct 2026 · github.com
Display limit
The current agent operates only the primary display, and multi-monitor automation is on the roadmap.Read 7 Oct 2026 · github.com
Installer formats
Build targets are Linux AppImage and deb, Windows NSIS, and macOS DMG.Read 7 Oct 2026 · github.com
Model compatibility
The model endpoint must support image inputs, function tools, and the model's tool-call format.Read 7 Oct 2026 · github.com
Model options
It can use a local vision-language model through vLLM or a compatible hosted API.Read 7 Oct 2026 · github.com
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.Read 7 Oct 2026 · github.com
Project maturity
The project describes itself as early alpha and warns that desktop actions should be supervised, especially around sensitive or irreversible workflows.Read 7 Oct 2026 · github.com
Security design
The Electron renderer is sandboxed with Node.js integration disabled, while privileged model and desktop automation behavior runs in the main process.Read 7 Oct 2026 · github.com
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.Read 7 Oct 2026 · github.com
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.Read 7 Oct 2026 · github.com
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.Read 7 Oct 2026 · github.com

Strong points and limits

Strong points

  • Local-first screen-aware desktop control
  • Supports local vLLM or hosted models
  • Ask-first approval is enabled by default
  • Open-source MIT license
  • Linux, Windows, and macOS build targets

Limits

  • Described as early alpha
  • Only the primary display is supported
  • Windows and macOS are experimental
  • API keys lack operating-system keychain storage

Who it suits

It suits users seeking local desktop automation on Linux, Windows, or macOS who can configure a compatible model endpoint. It is especially relevant to people who want approval controls before desktop actions.

Verdict

OpenComputer provides local, model-driven desktop control with explicit approval options and multiple installer formats. Its early-alpha status, experimental platforms, single-display limit, and model-endpoint requirements matter for deployment.

Where Specifiction places it

Alternatives to OpenComputerAll alternatives →