A production-ready AI assistant that breaks complex goals into sequential sub-tasks, executes them through direct device controls, and dynamically replans when conditions change.
Watch OpenDroid come to life with a friendly greeting, rendered in stunning 3D.
OpenDroid saying hi — rendered in 3D ✨
OpenDroid combines autonomous planning, deep device integration, and multi-modal AI into one seamless agent.
Breaks complex multi-step commands into logical sub-tasks, executes sequentially, and dynamically replans on failure.
Brightness, Wi-Fi, Bluetooth, flashlight, alarms, timers, calendar, calls, SMS — all controlled autonomously.
Clicks, scrolls, reads screens, and automates third-party apps like WhatsApp and Maps via Accessibility Services.
Captures and analyzes screenshots using multimodal LLMs. Falls back to text-scraping on older devices.
Working, episodic, semantic, and procedural memory systems for persistent context and user preference learning.
Hands-free wake word detection, speech recognition, and high-fidelity TTS with ElevenLabs fallback support.
A futuristic glassmorphic interface built with Jetpack Compose — dark mode, neon accents, and intelligent agent feedback.
Built with Dagger-Hilt dependency injection, clean architecture layers, and Jetpack Compose UI.
Services & app automators
System, comms, productivity
Planner, classifier, resolver
10+ provider integrations
4-tier persistent memory
Wake word, STT, TTS
Room DB, repos, models
Compose screens & themes
Download OpenDroid and give your phone the power of autonomous AI. Free, open source, and community driven.
