Computer use & desktop agents
727 agents · most starred first
Computer use & desktop agents, page 2 of 8
Automatic computer use agent for end-to-end automation workflows on macOS, Windows, Linux, iOS and the browser.
AI agent that controls computer with OS-level tools, MCP compatible, works with any model
EvoCUA: Evolving Computer Use Agent
Fazm Desktop for macOS
Fast computer automation CLI for AI agents. Control any desktop with screenshots, clicks, typing, scrolling, and more.
MoLing is a computer-use and browser-use based MCP server. It is a locally deployed, dependency-free office AI assistant.
Desktop automation MCP server — computer use for any AI agent: control screen, windows, mouse/keyboard, and Chrome via Model Context Protocol (stdio). Not a DSH plugin — DSH users install dsh-nuphus-mcp.
Superfast action layer for computer-use agents, on MacOS.
MCP-based Computer Use service for Qwen Code and any AI agent — controls macOS, Linux, and Windows via accessibility APIs.
An open-source implementation of Anthropic's Computer Use to perform basic tasks using AI Agents.
Free offline AI assistant for Windows — reads your screen, hears your question, speaks the answer. No API key, runs locally with Ollama.
[ICLR'25 Oral] UGround: Universal GUI Visual Grounding for GUI Agents
Human-taught Computer-use Agent Designed for Real Windows and MacOS Desktops.
PC Agent: While You Sleep, AI Works - A Cognitive Journey into Digital World
Get a guided tour for anything, right on your screen.
A curated list of of awesome UI agents resources, encompassing Web, App, OS, and beyond (continually updated)
Open desktop AI agent for Windows, macOS and Linux, written from scratch: its own agent engine, its own browser tooling and its own drawing engine that shows what it explains.
MacOS Demo for Claude Computer Use
Gemini 2.5 Computer Use + OpenAI Realtime API + Claude Code Multi-Agent Experiment
Cross-platform remote desktop for macOS and Windows with hardware acceleration, multi-monitor support, clipboard transfer for text, images, files and folders, and MCP integration for AI agents to connect to and control remote computers.
AI cursor assistant for macOS, Windows, Linux. Hold a hotkey, ask a question about what's on your screen, get an instant answer. Vision-capable AI overlay with OpenRouter, Anthropic, OpenAI, Gemini. Local-first, BYO key, no telemetry.
[ICML-2026] Official implementation of "SEAgent: Self-Evolving Computer Use Agent with Autonomous Learning from Experience"
Official implementation of GUI-R1 : A Generalist R1-Style Vision-Language Action Model For GUI Agents
Opensteer turns your computer into an AI agent workstation that you can control from anywhere.
QAA AirType 是一个轻量级的局域网输入板,让你可以通过手机端的输入法(如豆包语音输入)来实现电脑端的文字输入(与Agent对话)。
An AI Agent that interacts with Windows OS at GUI level.
The AI Agent That Lives On Your Desktop And Uses It Like You Do!
A codebase for "Language Models can Solve Computer Tasks"
DomoDomo is a 100% client-side, Agentic offline-first web utility application. Built as a high-performance, private, zero-server architecture toolbox, all operations run completely inside your browser sandbox. Your data, images, PDFs, and files never leave your computer no servers, no APIs, and no external clouds are ever touched.
GitHub page for "Large Language Model-Brained GUI Agents: A Survey"
One command for every interface—search, run, and inspect real software across APIs, browsers, desktops, local tools, and MCP.
Isolated Linux desktop workspaces for AI agents — a hidden, agent-owned desktop and browser over MCP, so an agent can do GUI and web work without touching your real desktop.
A desktop for AI agents
An implementation of a computer use agent (CUA) using LangGraph
DeepSeek Harness 的桌面自动化:给整台 Windows 机器装上眼睛和手。可以做人能做的任何事,包括且不仅限于CAD,WPS,浏览器操作,发微信。安全,快速,可靠,聪明,人性化。
Goodable is a local-first Desktop AI Workspace (Desktop Agent Runtime) powered by Claude Agent SDK, combining OS/file control, browser automation, and coding into composable Skills (tool + app modes) to build, run, and publish real AI apps.
A toolkit for building computer use AI agents
[ACL 2025] Code and data for OS-Genesis: Automating GUI Agent Trajectory Construction via Reverse Task Synthesis
Computer-using agent built with Next.js and the AI SDK by Vercel.
AI controls your OS. OS AI Computer Use, OS and API agnostic. For now on OpenAI and Anthropic API. Desktop app ready.
System 1 decision models (Jev, Laya, Cua-S1) as brain for agents: Browser use, computer use, games and robotics
Spongecake is the easiest way to launch computer use agents.
An AI companion that lives on your desktop - OpenGuider watches your screen, listens to your voice, and guides every step with clear actions.
Teach AI how your work actually gets done. Capture real computer work, extract the workflow graph and judgment, then run it with Codex or your own agent.
A native macOS assistant that reads the current screen and browser context when you need it.
Welcome to **LiveRecall**, the open-source alternative to Microsoft's Recall. LiveRecall captures snapshots of your screen and allows you to recall them using natural language queries, leveraging semantic search technology. For added security, all images are encrypted.
An Multi-Modal LLM Powered Agent to automatically solve Captchas
Autonomous desktop intelligence with MCU Ultron persona, mood-aware workspace orchestration, cyber security sentry, and full PC control.
docker region:us
Give any AI agent a full desktop — it sees the screen, clicks, types, and runs apps like a human. Automate anything with a UI: browsers, legacy software, internal tools. No API needed. One Docker command.
A multi-agent collaboration framework that lets AI agents work together, operate your computer, and browse the web.
Open-source alternative to Claude Co-Work built by Claude. with BYOK
Moved to github.com/davidondrej/cloudroom
基于多模态视觉感知与 LLM Agent 的 macOS 微信自动化框架 | Visual RPA for WeChat
An open-source take on Grok Bot/Muse. Persistent AI teammates with their own cloud computers, powered by Runta.
MCP server for computer use & browser automation - screenshot, OCR, click, type, find_text, Chrome/Electron CDP, template matching. macOS, Windows & Android. Works with Claude, Cursor, and any MCP client.
If it's on the screen, it's an API. Control any desktop app via REST + MCP. Rust.
App-Controller: Allow users to manipulate your App with natural language
GPT-4V in Wonderland: LMMs as Smartphone Agents
Small enough to run on your Mac. Smart enough to ask. An open-source computer use agent for macOS, with local models (MLX) that see, decide and act. 得心,应手。
Say it, and your Mac does it. A computer-use harness on Jev that reads the screen through Accessibility. Fast, no vision model
Turn intent into work. Open-source Desktop/Web agent harness with long-horizon memory, live evaluation and recoverable batches.
Enable AI to control your PC. This repo includes the WorldGUI Benchmark and GUI-Thinker Agent Framework.
[NeurIPS 2025] Official repository of RiOSWorld: Benchmarking the Risk of Multimodal Computer-Use Agents
MCP server for computer use & browser automation - screenshot, OCR, click, type, find_text, Chrome/Electron CDP, template matching. macOS, Windows & Android. Works with Claude, Cursor, and any MCP client.
TUI automation for AI agents. Control any terminal app from code.
CaoGen — a vendor-neutral, local-first AI work desktop. Multi-provider agent runtime, GUI automation, MCP, Git worktree, task orchestration, effect ledger, and 3D office workspace. macOS/Windows/Linux.
An open-source agentic framework that enables AI to use computers like humans and can provide a multi-agent runtime environment as an infrastructure capability
Auto-Use Computer Use — drives your OS, browser, scours the web, writes your code. One agent, end to end.
A lightweight, blazing-fast AI agent for Windows. Automate your OS via CLI control, built-in tools, MCP tools, custom Python scripts, and expandable skills. Fully autonomous.
CaoGen — a vendor-neutral, local-first AI work desktop. Multi-provider agent runtime, GUI automation, MCP, Git worktree, task orchestration, effect ledger, and 3D office workspace. macOS/Windows/Linux.
InterfaceAgent: a versatile framework designed to create system and interface agents capable of managing mobile and desktop applications and features.
CaoGen — a vendor-neutral, local-first AI work desktop. Multi-provider agent runtime, GUI automation, MCP, Git worktree, task orchestration, effect ledger, and 3D office workspace. macOS/Windows/Linux.
Turn Android Termux into an autonomous AI agent and security orchestrator. Powered by a local ReAct framework with 50 system tools and MCP support, a background task │ scheduler, on-device ADB controls, and a unified Web UI & Telegram Bot.
Your Apple-Native AI Agent
Let agents control any remote computer, anytime, anywhere. Self-hosted AI agent remote assistance via MCP and terminal.
A free, open-source AI agent that runs entirely on your machine — controls your computer like you would, works across any app, and handles arbitrary tasks: writing code, analyzing images, organizing files, researching and building spreadsheets, whatever you throw at it. Unlike Claude Cowork, nothing leaves your device.
On-device computer use agent that runs fully in the background 🪄
Free open-source Screenpipe GUI + macOS automation agent. Search your screen history, view analytics, automate workflows — all local, no cloud. Works with the free Screenpipe CLI.
IRIS 👁️ — A "God Mode" AI operator that controls your entire OS 💻 via voice. Automates coding 👨💻, deep focus 🛡️, remote access 📲, and context recall ⏳. No mouse, no keyboard → complete intelligent execution 🚀.
Nuphus — 本地优先的 AI Agent:真实桌面执行力 + 手机第二块屏幕。Local-first AI agent with real desktop execution and dual-device real-time sync.
The setup I used for my GPT-6 Astra Portal run: controller, SourcePauseTool patch, setup guide, and sanitized session log.
Computer use for the Pi coding agent — Windows apps, browsers and desktops through native UI trees, screenshots and real input.
Open Agent Computer Interface
Local hands for your AI, keys in your pocket — a desktop agent where every capability is authorized, executed locally, and backed by Evidence.
CUA-Lite: An Open Platform for Computer-Use Agents
Desktop agent framework for creating AI agents that can see and control your computer through voice and text commands
DART-GUI: Efficient Multi-turn RL for GUI Agents via Decoupled Training and Adaptive Data Curation
Let AI agents control Windows applications. Click buttons, type text, toggle settings — all by name, not coordinates. Uses the Windows UI Automation API to find UI elements reliably, regardless of DPI, theme, resolution, or window position.
Let agents control many desktop software directly from the cli, with one pip install, and no MCP servers.
Let your AI apps use your Mac: a local server and menu bar app that lets Claude, ChatGPT, Codex and other MCP apps work in your Mac's apps, files, terminal and browser, with approvals you control.
Official repository for SaaS-Bench: realistic, locally deployable SaaS workflows for GUI agent evaluation.
Native VNC, RDP and SSH client for Windows, macOS and Linux. Pure-Rust protocol cores, Tauri 2. One host library for all three, plus an MCP control plane so an AI agent can drive remote desktops with human takeover.
An open source environment for digital agents.
Copper (Cu) CoAP user-agent (JavaScript implementation)
A Swift library for macOS automation — mouse, keyboard, screenshots, image recognition, and AI-powered agents.
An open-source visibility and management layer for AI tools across a desktop fleet
Official code for "Focused Omni-Modal GUI Agents“
Your computer was paying attention the whole time. 14-modality capture. Proactive intelligence. Computer-use training data. Native macOS. All on-device. Open source.
touri: open-source desktop app for AI agent bots (Codex, Claude, OpenCode) with isolated VMs, browser and computer use.