A centralized, privacy-first AI platform for mobile — download models once, use everywhere. One app to manage all your AI models, with inter-app sharing so other apps don't need to bundle their own.
Today's mobile AI ecosystem is wasteful and inefficient.
Every AI app downloads its own models (100MB-4GB each). 5 apps with the same model = 5 copies wasting storage.
Multiple apps running separate inference engines destroys battery life and heats up your device.
Apps can't share AI capabilities. Each reinvents inference, chat history, and tooling from scratch.
Download models once, share with all apps, keep your data private.
Download GGUF models from Hugging Face once. They're available to all compatible apps via a local API.
ReadyOther apps can POST to localhost:8080 for AI inference. No SDK needed — OpenAI-compatible API.
PreviewAll inference runs on-device. Your data never leaves your phone. No telemetry, no cloud uploads.
ReadyScreenshots from the actual app — privacy-respecting productivity at your fingertips.
Personalized greeting with daily digest. Quick access to AI status, tasks, and upcoming events.
home_screen.dart
Real-time streaming chat with function calling. AI executes tasks, logs expenses, and searches notes.
chat_screen.dart
Notes with search bar and tag filters. FTS5 full-text search for instant retrieval.
brain_screen.dart
Inline editing with tags. Quick capture and organize knowledge for AI-powered retrieval.
brain_screen.dart
Drag-and-drop task board. Todo, In Progress, Done — visual workflow management.
tasks_sub_screen.dart
Expense and income tracking. AI can log transactions via natural language.
finance_sub_screen.dart
All available AI capabilities: add_task, log_expense, search_notes, get_weather, and more.
tools_sub_screen.dart
Fully offline conversations powered by on-device GGUF models. No cloud, no data leaves your phone.
chat_screen.dart
Local HTTP server for other apps. POST to localhost:8090 — OpenAI-compatible API.
gateway_sub_screen.dart
Swagger-like UI served at localhost. Authenticate with access code, test endpoints with live streaming.
api_playground_html.dart
Configure AI providers, themes, privacy settings. Export/import your data.
settings_screen.dartChat, tools, knowledge base, and more — all in one unified interface.
Chat with AI models using LangChain.dart. Token-by-token streaming, conversation history, and model switching mid-chat.
ReadyLocal GGUF models (Gemma, Llama, Phi), Ollama, OpenAI, Gemini, OpenRouter, Hugging Face — all seamlessly switchable.
ReadyAI can execute tools: add tasks, log expenses, search notes, create notes, and get weather. Full LangChain ToolSpec integration.
ReadyAreas → Resources → Notes hierarchy with FTS5 search. Custom personas and a Soul document for AI personalization.
ReadyManage tasks with priority levels and due dates. Track expenses and income with category-based organization.
ReadyOCR for text extraction, entity extraction for parsing, smart reply suggestions, and language identification — all offline.
ReadyLocal-first with reactive streams. 10 tables for conversations, messages, tasks, transactions, areas, resources, notes, and more.
Ready7 accent color presets, AMOLED dark mode, and responsive design. Built with Moon Design components for consistency.
ReadySpeech-to-text for hands-free interaction. Full integration with speech_to_text package for natural conversation.
ReadyEphemeral conversations that don't save to history. Perfect for quick questions you don't need to keep.
ReadyNo telemetry, no cloud uploads, no tracking. Everything runs and stores locally by default.
All data stored on your device using Drift SQLite. No cloud uploads unless you explicitly choose a cloud AI provider.
Use Ollama or llama.cpp for true local inference. Your conversations never leave your phone.
Zero analytics, zero tracking, zero server calls. Open source and fully auditable.
AGPL-3.0 licensed. Audit the code, build from source, contribute. Full transparency.
Track our progress from initial UI to a fully-featured AI assistant.
Full app shell, themed UI, real-time streaming chat, tool calling (14 tools), Drift database, local GGUF inference, cloud providers, API Gateway with Swagger-like playground, model catalog with HuggingFace downloads.
Full voice input/output, retrieval-augmented generation (RAG) for knowledge base queries, and smart proactive suggestions.
Calendar sync (Google/Apple), wake word detection ("Hey Prism"), continuous listening, and notification actions.
Home screen widgets, encrypted multi-device sync, plugin system for third-party extensions, iOS support.
Stop duplicating models across apps. Download v1.0, explore the source, and experience centralized offline AI.