HomeProjectsArticlesPK BoardAbout
Explore Projects

MobileAgent

Normal
GitHub Python MIT

Description

Mobile-Agent is the GUI agent family from Alibaba Tongyi Lab, specializing in natural-language control of Android and iOS devices.

Key Features

  • GUI agent family — Complete Mobile Agent lineup from Tongyi Lab
  • Multi-platform — Covers both Android and iOS device operations
  • Visual perception — Understands mobile UI state via screenshots and OCR
  • Multi-step planning — Decomposes complex tasks into executable step sequences
  • Model variety — Multiple model versions including Mobile-Agent-v3.5
  • Online demos — Available on Modelscope and Alibaba Bailian

Use Cases

💡 Automating app operations on Android and iOS devices
💡 Providing mobile operation assistance for visually impaired users
💡 Automating clicks and swipes for testing scenarios
💡 Building cross-platform mobile RPA workflows
💡 Benchmarking different GUI agents on mobile tasks

Strengths & Limitations

✅ Strengths

  • • High community interest (9.1k stars)
  • • Permissive open-source license (MIT)
  • • Established track record (2 years in production)

Tags

mobile gui-agent computer-use vision android ios

Categories

🌐 Browser Agent ⚡ Agent Tools

Quick Start

# Clone the repo
git clone https://github.com/X-PLUG/MobileAgent.git
cd MobileAgent

# Install dependencies
pip install -r requirements.txt

# Run a demo script
python demo_android.py
Visit GitHub Visit Website

Related Projects

Mobilerun

9.1k · Python
Active A+

Mobilerun is an open-source Android and iOS device automation framework that lets LLM agents control mobile devices via screenshots and UI state.

mobile-automationandroidios +3
  • · LLM-agnostic — Compatible with any LLM provider, switch models on demand
  • · Mobile-native tools — Screenshots, UI state, tap, swipe and type as agent tools
  • · Multi-step planning — Agents autonomously plan and execute multi-step tasks
View full details → View on GitHub

Agent S

12.2k · Python
Active A+

Open-source agentic framework that uses computers like a human, capable of completing complex GUI tasks with autonomous learning and experience accumulation.

computer-usegui-agentautomation +1
  • · First framework to surpass human-level performance on OSWorld (72.60%) benchmark
  • · Cross-platform support for Linux, macOS, and Windows with single-monitor GUI interaction
  • · Modular grounding with UI-TARS-1.5-7B and support for multiple LLM backends (OpenAI, Anthropic, Gemini)
View full details → View on GitHub

Agent Device

4.1k · TypeScript
Active A

CLI to control iOS and Android devices for AI agents, enabling coding agents to directly interact with mobile devices for testing and automation.

mobileiosandroid +3
  • · Device automation CLI for iOS, Android, TV, and desktop app verification by AI agents
  • · Token-efficient UI snapshots with semantic refs like @e1, @e2 for element interaction
  • · Evidence capture including screenshots, videos, logs, traces, network traffic, and crash context
View on GitHub