Welcome to DroidRun

DroidRun is a powerful framework that enables you to control Android devices through LLM agents. It provides a simple and intuitive way to automate Android device interactions using natural language commands.

Features

Natural Language Control

Control your Android device using natural language commands

Multiple LLM Support

Support for OpenAI, Anthropic, Gemini, Ollama, and Deepseek

Advanced Planning

Optional planning and reasoning capabilities

Vision Support

Built-in vision capabilities for screen analysis

Simple CLI

Rich terminal UI with live updates

Python SDK

Comprehensive SDK for custom automation tasks

Quick Example

# Simple CLI usage
droidrun "Open the settings app"

# With specific provider and model
droidrun "Open calculator app" --provider Gemini --model models/gemini-2.5-pro-preview-05-06

# With vision and planning capabilities
droidrun "Open Calculator and take a screenshot" --vision --reasoning

Or with Python:

import asyncio
from droidrun.agent.droid import DroidAgent
from droidrun.agent.utils.llm_picker import load_llm
from droidrun.tools import load_tools

async def main():
    # Load tools and LLM
    tool_list, tools_instance = await load_tools()
    llm = load_llm(
        provider_name="Gemini",  # OpenAI, ollama, Anthropic, Gemini, DeepSeek
        model="models/gemini-2.5-pro-preview-05-06",
        temperature=0.2
    )
    
    # Create and run the agent
    agent = DroidAgent(
        goal="Open the Settings app",
        llm=llm,
        tools_instance=tools_instance,
        tool_list=tool_list,
        vision=True,      # Enable vision capabilities
        reasoning=True,   # Enable planning mode
        enable_tracing=True  # Enable execution tracing
    )
    
    result = await agent.run()
    print(f"Success: {result['success']}")
    if result.get('output'):
        print(f"Output: {result['output']}")

if __name__ == "__main__":
    asyncio.run(main())

Prerequisites

  • Android device connected via USB or ADB over TCP/IP
  • ADB (Android Debug Bridge) installed
  • API key for at least one LLM provider
  • DroidRun Portal app installed and accessibility service enabled

Getting Started

Core Concepts