AI · Website

ACU - Awesome Agents for Computer Use

Curated collection of papers, tools, and frameworks for computer-controlling AI agents

Visit github.com ↗

ACU is an open-source knowledge base tracking the ecosystem of AI agents that operate computers and mobile devices autonomously via clicks, keystrokes, and APIs. It organizes recent research papers, UI grounding techniques, evaluation benchmarks, and both open-source and commercial frameworks. It serves as a central starting index for developers and researchers building device-control agents.

What it's for

  • Discover open-source frameworks and environments for building computer-use agents
  • Track research papers across GUI grounding, agent safety, and trajectory synthesis
  • Find evaluation benchmarks and datasets for measuring desktop and mobile agent performance

Features

  • Taxonomy covering surveys, foundation models, UI grounding, datasets, benchmarks, and safety
  • Indexed list of open-source frameworks including Browser Use, OpenInterpreter, and AutoGen
  • Curated links to benchmark environments like OSWorld, AndroidWorld, and Windows Agent Arena
  • Coverage of UI grounding models such as OmniParser and Ferret-UI
  • Community contributions tracked via GitHub

Pricing

Free

Details

Type
Website
Category
AI
Topic
Curated Lists
Open source
Yes
Platforms
Web
For
AI researchers, software engineers, and robotic process automation developers
Tags
computer use, ai agents, gui agent, llm, automation, awesome list

Why it was saved

  • @omarsar0, 7 February 2025: One comprehensive starting point for anyone building or researching computer-use agents: an indexed collection of papers, projects, frameworks and tools rather than scattered sources.

Related

  • Ace

    Realtime desktop autopilot that controls mouse and keyboard for native apps.

  • Agent S

    Open-source computer use agent operating GUI interfaces via mouse and keyboard

  • OmniParser

    Parses UI screenshots into structured interactive elements for vision LLMs

  • Google Opal

    No-code visual canvas for building AI mini-apps from plain English prompts

  • Hermes Agent

    Open-source, self-hosted AI agent with persistent memory and messaging integrations.

  • AutoGPT

    Open-source platform and framework for building autonomous AI agents

Updated