Open AGI Apps

Practical applications demonstrating open artificial intelligence

Featured Applications

AGI Chat

Multi-Modal AI Interaction Hub

Open Source / Available
AGI Chat Screenshot

A versatile web-based AI chat interface. Connect to multiple models (DeepSeek, Gemini, SambaNova, free Proxy option), customize AI personas with unique prompts and profile pictures, and use the unique 'Helper AI' system to preprocess your messages (e.g., for translation or complex instructions). Integrates AI Search and Deep Dive tools, supports voice input, and saves chat history locally.

Key Features:

  • Multiple AI provider/model support + Free Proxy
  • Customizable AGI Personas (Name, Prompt, PFP Gen)
  • Helper AI for message preprocessing
  • Integrated AI Search & Deep Dive tools
  • Voice Transcription via Mic
  • Persistent chat history & management
  • Open Source (MIT Licensed)

AGI Companion Browser

AI Agent Browser for Android

Open Source / Available
AGI Companion Browser Screenshot

Control your Android web browsing visually using AI! AGI Companion Browser acts as an AI Agent, allowing advanced multimodal models (starting with Gemini, with Qwen, DeepSeek, Llama planned) to understand web pages via screenshots and interact using precise grid-based commands. Navigate, click, type, and automate tasks just by describing your goal. Explore the future of web interaction!

Key Features:

  • AI-powered visual web control (Gemini, others planned)
  • Grid-based targeting for clicks & typing
  • Multi-turn task execution & context memory via notes
  • Full browser features (tabs, desktop mode, downloads)
  • Configurable API Key support
  • Open Source (MIT Licensed)

Arcane Tome Chat

Themed AI Chat Interface

Available
Arcane Tome Chat Screenshot

Engage with AI models through a highly customizable, themed chat interface. Arcane Tome supports multiple API providers and models, persistent chat history, custom system prompts ("Guidance"), and various visual themes. Uniquely, settings and API/chat management menus are accessed by clicking the spirit (bot) and user profile pictures (sigils) respectively, providing a clean and immersive experience.

Key Features:

  • Multiple visual themes (Fantasy, Cyberpunk, etc.)
  • Multi-language support
  • Support for various API providers & models (configurable)
  • Saved chat history (via IndexedDB)
  • Customizable system prompts ("Spirit Guidance")
  • Avatar customization
  • Markdown rendering for AI responses
  • Unique avatar-click menu access
  • Open Source (MIT Licensed)

Sentinel AI Assistant

Multimodal Android AI Assistant

Open Source / Available
Sentinel AI Assistant Screenshot

An Android app providing a multimodal AI interface (initially Gemini). Interact with text and images, configure automatic image capture with custom prompts, and set up alert triggers based on AI responses, optionally initiating a phone call.

Key Features:

  • Multimodal interaction (Text & Image) via Gemini
  • Configurable API Key support
  • Auto-capture mode with custom intervals & prompts
  • Alert triggers based on AI response keywords
  • Optional phone call initiation on alert
  • Persistent settings via local storage
  • Open Source (MIT Licensed)

Development Roadmap

AGI Chat: Helper Enhancements

Expand the Helper AI system with more built-in preprocessing options and explore agent-like capabilities for the helpers themselves.

Browser: Multi-Model Integration

Integrate additional multimodal models like Qwen, DeepSeek, and Llama into the AGI Companion Browser for broader compatibility and choice.

Arcane Chat: Enhanced Tool Use

Explore integrating function calling or tool use capabilities within the Arcane Tome chat interface for more complex interactions.

General: Plugin Ecosystem

Expansion of the plugin system to support community-developed extensions for specialized tasks and integrations across apps.

General: Advanced Features

Implementation of more advanced AGI capabilities including long-term memory, task automation, and self-improvement features.