The request names "open-computer-use", which none of the sources identifies as a product, so no single tool is confirmed as that name. For giving an AI agent computer use (screen, mouse, keyboard) the closest matches are Orgo, which provides cloud desktops agents control 2, Glasswarp, a computer-use API for real Windows PCs 1, Browser Use for browser-level agent control 3, and agent-device, an open-source device layer exposed via CLI and MCP 4.
A question back: Did you mean a specific product by "open-computer-use", or do you want a Windows, Linux, or macOS desktop, and local or cloud?
- 1
Orgoorgo.aiverified and signed
Provides secure cloud Linux or Windows desktops where agents see the screen and use mouse and keyboard, with actions to control a computer and run agent tasks 2.
+1 community tools built for it
for agents: Control Computer · Run Agent Task · Provision Computer
- 2
Glasswarpglasswarp.comwebsite read, tools not yet checked
Gives any model eyes and hands on a Windows PC you already own, driven from Python, Cursor, or MCP after one-time pairing 1.
official:MCP server · Glasswarp+1 community tools built for it
- 3
Browser Usebrowser-use.comverified and signed
Cloud browser control for AI agents, covering web-level interaction, automation, and screenshots rather than full desktop control 3.
official:MCP server · com.browser-use/browser-use
for agents: browser-use · open-source · browser-harness
- 4
Sources
- 1Glasswarp — Computer use API for real Windows.glasswarp.com · read 2026-10-29
- 2Orgo - Computers for AI agentsorgo.ai · read 2026-10-29
- 3Browser Use Agents & Browser Infrastructure | Browser Usebrowser-use.com · read 2026-10-29
- 4agent-device: Open-source device layer for agentsagent-device.dev · read 2026-10-29
For agents: the 10 actions behind this answer
· Glasswarp website read, tools not yet checked
Computer use API for real Windows. Give any model eyes and hands on a Windows PC you or your client already own — not a rented sandbox. Pair once; drive from Python, Cursor, or MCP.
Control Computer · Orgo Agent verified and signed
Drive a running computer: click, type, scroll, take screenshots, and run bash or Python. e.g. Open Chrome and screenshot the orgo.ai homepage
browser-use · browser-use verified and signed
Direct browser control via CDP for web interaction: automation, scraping, testing, screenshots, and site/app work. Also covers hosted cloud browsers for AI agents: the browser-use Python library (Browser(use_cloud=True)), Playwright, Puppeteer or any CDP client.
· agent-device website read, tools not yet checked
Open-source device layer for coding agents: control mobile, web, TV and desktop apps on devices and simulators, local or in the cloud, via CLI, MCP or Node.js.
ssh · ssh verified and signed
Establish and manage SSH connections to remote machines, including key generation, configuration, and file transfers. Use when connecting to remote servers, executing remote commands, or transferring files via SCP.
· Strands Agents website read, tools not yet checked
Strands is the open source toolkit for building production AI agents in Python and TypeScript: an agent harness and SDK, sandboxed shell, evals, and research labs.
· Clawdcursor website read, tools not yet checked
Local MCP server that gives any AI agent safe desktop control. Compiles the screen into a UI map (a11y tree + OCR) with stable element ids, acts by id not by pixel, verifies its own actions, routes everything through one safety gate. Windows · macOS · Linux. No cloud.
livekit-simulations · livekit-simulations verified and signed
Generate targeted test scenarios for a LiveKit voice or chat agent and run them as simulations — locally, from the agent's own code plus what the user wants stress-tested. Use whenever the user wants to "test my agent", "what should I test", "create/generate simulation scenarios", "make a sim test s
automation · robotframework verified and signed
· Herdr: the runtime coding agents run on website read, tools not yet checked
Herdr is the runtime your coding agents live on. It holds real terminals open on your laptop, your desktop or a box you rent, so the work survives the lid closing — and you can attach again from anything with a keyboard.
Labels are the instruments' verdict and are never changed by ranking. Agents and robots get this same answer as JSON at /api/v1/find?q=…&answer=1 and through the MCP server.