AIエージェント
UI-TARS-desktop
ByteDance's open-source multimodal AI agent stack: GUI agents that see and operate real desktops
★ 39.2k スターApache-2.0
ガイドを見る →MobileAgentは、X-PLUGが公開しているオープンソースのAIエージェントプロジェクトです（GitHub: X-PLUG/MobileAgent）。GitHubスターは9,281、フォークは930で、MITライセンスの下で公開されています。GitHubでの最終更新は2026年7月7日です。
Mobile-Agent — multimodal GUI agents that operate phones and PCs from screenshots and instructions.
Pick which member of the Mobile-Agent family and which scenario (phone, PC or web) fits your task, and check its specific setup in the repo.
Clone X-PLUG/MobileAgent from GitHub (or use the install path in the run steps below) and skim its README for the parts relevant to your goal.
Clone the repo and follow the chosen agent's instructions for dependencies, device connection and model configuration.
Run one of the provided demos on a connected device or emulator with a simple instruction, and review the agent's step-by-step actions.
The repo hosts the Mobile-Agent series of works, so you can pick the variant that matches your device and task type.
Agents are built on multimodal LLMs that read screenshots and screen content, not just view hierarchies.
Designed to operate real smartphone apps end to end from natural-language instructions.
Demos and variants cover desktop PC and web interfaces as well as phones.
The project publishes benchmark results for its agents alongside the code and demos.
The full family is available under the MIT license for research and adaptation.
Get the source from the official GitHub repository for MobileAgent.
git clone https://github.com/X-PLUG/MobileAgent.git
cd MobileAgentInstall the requirements for the Mobile-Agent variant you chose, following that variant's section in the README.
pip install -r requirements.txt # see README for the exact variant's requirements fileConnect a device or emulator, configure the model per the README, and launch a demo task.
# Follow the demo instructions for your chosen Mobile-Agent variant — see READMEX-PLUG, through the X-PLUG/MobileAgent repository.
Yes — the project is open source under the MIT license as recorded for this listing. You can use, modify and distribute it under the MIT terms.
The agents are driven by multimodal LLMs, so you typically configure your own model access or API key per the variant's instructions; the code itself is free.
Scripts replay fixed taps and break when apps change. MobileAgent's models perceive the live screen and decide actions dynamically — more general, but a research codebase where each variant has its own setup, rather than a packaged product.
Clone the repository (or use the documented installer), follow the install and configure steps on this page, and run the smallest real task first — run one of the provided demos on a connected device or emulator with a simple instruction, and review the agent's step-by-step actions.
ByteDance's open-source multimodal AI agent stack: GUI agents that see and operate real desktops
Microsoft's GUI agent that grew into UFO³ — orchestrating agents across Windows, Linux and mobile devices.
Python library for building multimodal language agents — VLM, video and mobile, fast to prototype and ship
Android GUI agent framework: AI that sees, plans, and taps real mobile apps.
Mobile app automation and verification built for AI coding agents — CLI, MCP server, and Node API
Out-of-the-box open-source AI agent for desktop and mobile. Just works.
LibreAIカタログに掲載されたAIエージェントプロジェクトの中で最もスターの多いもの — MobileAgentが合わなかった場合に検討したい、現実的な候補リストです。
A personal AI that really does things — on any OS, with your data staying yours.
An agentic skills framework & software development methodology that works.
Agent harness optimization system with skills, memory and security for Claude Code, Codex and more
The agent that grows with you — a Python framework for adaptive AI agents.
DeepSeek's plugin-based agent harness where everything is a plugin.
カタログの中でMobileAgentとスター数が近いAIエージェントプロジェクト — 前後の棚もそのまま見て回れます。
Scan your LLM for vulnerabilities before attackers do.
Deep research that stays on your machine — local or cloud LLMs, 10+ search engines, encrypted and private.
The open runtime for AI agents — a Rust platform for running, governing and routing agents with TokenOps.
A GEP-powered self-evolving engine for AI agents — auditable evolution with Genes, Capsules and Events.
Control Android and iOS devices with natural language — an LLM-agnostic mobile agent framework and CLI.
PraisonAI 🦞 — Hire a 24/7 AI Workforce. Stop writing boilerplate and start shipping autonomous self-improving agents that research, plan, code, and execute t…