OpenDroid

OpenDroid isn't just another chatbot. It's a fully autonomous AI agent that lives on your Android phone and actually does things for you.

"Check if it's going to rain tomorrow, and if so, text my wife that I'll be late and set an alarm for 6 PM."

OpenDroid will plan this as 3 steps, execute each one, verify the results, and adapt if anything fails — all without you lifting a finger.

✨ Features

- Autonomous Agent Engine 🧠 Capability: Self-Planning Description: Breaks complex commands into sequential steps with dependency tracking

- Autonomous Agent Engine 🧠 Capability: Re-Evaluation Description: Monitors execution results and dynamically replans when steps fail

- Autonomous Agent Engine 🧠 Capability: Compound Intent Guard Description: Smart detection of multi-action commands (e.g. "open WhatsApp and send message")

- Autonomous Agent Engine 🧠 Capability: Contact Disambiguation Description: 4-tier contact resolution with fuzzy matching and relationship aliases ("call dad")

🛠️ On-Device Model Manager (LiteRT-LM)

- Background Downloader Capability: Background Downloader Description: Real network downloads (via WorkManager) with Pause/Resume, speed tracking, and ETA

- Secure Authentication Capability: Secure Authentication Description: Encrypted token storage (EncryptedSharedPreferences) to securely fetch gated Hugging Face models

- Integrity Verification Capability: Integrity Verification Description: Computes SHA-256 hashes and verifies LiteRT engine loading compatibility before marking READY

- Local Model Import Capability: Local Model Import Description: Direct offline importing of custom .task or .litertlm files with JNI verification checks

📱 Full Device Control

- System Action: System Examples: Brightness, WiFi, Bluetooth, Flashlight, DND, Volume, Screenshot

- Communication Action: Communication Examples: Calls, SMS, WhatsApp messages, Email drafts

- Productivity Action: Productivity Examples: Alarms, Timers, Reminders, Calendar events, Notes

- Navigation Action: Navigation Examples: Google Maps directions, Uber/Ola booking

- Media Action: Media Examples: Play/pause music, YouTube search, camera

- Finance Action: Finance Examples: UPI payments, bill splitting, currency conversion

- Smart Home Action: Smart Home Examples: Google Home device control

👁️ Vision Engine

Captures screenshots via Accessibility API and feeds them to vision-capable LLMs for real-time screen analysis. Falls back to accessibility tree text-scraping on older devices.

🗄️ Multi-Tier Memory System

┌─────────────────────────────────────────────────┐ │ Memory System │ ├──────────────┬──────────────┬───────────────────┤ │ Working │ Episodic │ Semantic │ │ (current │ (past task │ (long-term facts │ │ context) │ results) │ & preferences) │ ├──────────────┴──────────────┴───────────────────┤ │ Procedural Memory │ │ (user-defined macro workflows) │ └─────────────────────────────────────────────────┘

🎙️ Voice Interface

- Offline wake word detection — say "OpenDroid" to activate - Speech-to-text for hands-free commands - Text-to-speech with ElevenLabs premium voice support

🎨 Premium UI

Built with Jetpack Compose featuring a futuristic glassmorphic design:

- Deep navy (#080C10) + Neon green (#00FF88) color system - Pulsing audio orb animation during listening - Live latency benchmarks for each provider - Dark mode by default

🔌 Supported LLM Providers

OpenDroid supports 12 LLM providers with automatic failover:

- Google Gemini 🟢 Provider: Google Gemini Models: Gemini 2.0 Flash, Pro, Nano Type: Cloud + On-device

- Anthropic Claude 🟣 Provider: Anthropic Claude Models: Claude Sonnet 4, Opus 4 Type: Cloud

- OpenAI 🔵 Provider: OpenAI Models: GPT-4o, GPT-4.1, o3 Type: Cloud

- Groq ⚡ Provider: Groq Models: LLaMA 3, Mixtral (ultra-fast) Type: Cloud

- DeepSeek 🔷 Provider: DeepSeek Models: DeepSeek V3, R1 Type: Cloud

- Mistral AI 🟠 Provider: Mistral AI Models: Mistral Large, Medium Type: Cloud

- OpenRouter 🌐 Provider: OpenRouter Models: 200+ models via unified API Type: Cloud

- Together AI 🤝 Provider: Together AI Models: Open-source model hosting Type: Cloud

- Cohere 🔴 Provider: Cohere Models: Command R+ Type: Cloud

- GitHub Copilot 🐙 Provider: GitHub Copilot Models: GPT-4.1, Claude via Copilot API Type: Cloud

- Ollama 🏠 Provider: Ollama Models: Any local model (LLaMA, Phi, etc.) Type: Local

- Custom OpenAI 🔧 Provider: Custom OpenAI Models: Any OpenAI-compatible endpoint Type: Self-hosted

Required Permissions

On first launch, OpenDroid will guide you through granting:

- Accessibility Service 🔓 Permission: Accessibility Service Why: UI automation, screen reading, app control

- Write Settings ⚙️ Permission: Write Settings Why: Toggle WiFi, Bluetooth, brightness

- Record Audio 🎤 Permission: Record Audio Why: Wake word detection & voice commands

- Notification Access 🔔 Permission: Notification Access Why: Smart notification reading & auto-reply

- Post Notifications 📱 Permission: Post Notifications Why: Foreground service status

📜 License

Copyright 2026 OpenDroid Contributors

Apache 2.0