DevIO: A Local-First Flutter AI Chat Client That Balances Privacy and Intelligence

DevIO is a cross-platform mobile app built with Flutter, centered on the core concept of "local-first". It allows users to run large language models on their devices (via tools like Ollama and LM Studio) for fully offline AI chat experiences. Its core value lies in protecting user privacy, eliminating network dependency, and supporting a multi-provider architecture—users can freely choose backend services, balancing intelligence with data control.

Most current AI interactions rely on cloud APIs, which pose privacy leakage risks and network dependency issues. DevIO proposes the idea of "bringing AI capabilities back to local devices" to address these pain points, making it an ideal choice for privacy-conscious users and developers.

• Ollama Provider: Default local option, using streaming APIs to support offline open-source models (e.g., Llama, Mistral);
• LM Studio Provider: Integrated via OpenAI-compatible API, requiring a local server to be started;
• OpenAI-Compatible Provider: Supports any compliant inference service;
• Anthropic Provider: Extends support for Claude models. Users can freely choose backends without being locked into a specific platform.

• Data Persistence: Uses Drift/SQLite to store chat history, with database files kept on local devices;
• Data Migration: Automatically migrates data from older SharedPreferences versions to SQLite for seamless upgrades;
• State Management: Uses BLoC/Cubit to separate business logic from UI, facilitating testing and maintenance.

• Streaming Chat: Real-time display of tokens generated by the model to enhance interaction fluency;
• Chat Management: Supports creating, pinning, renaming, deleting chats, and searching history;
• Demo Mode: Experience app features without an LLM server;
• Personalized Settings: Configure default providers, server addresses, and other parameters.

• Privacy-Sensitive Scenarios: Lawyers, doctors, etc., can ensure data never leaves their local devices;
• Network-Restricted Environments: Offline availability, unaffected by network conditions;
• Developers/Researchers: Quickly test different model checkpoints;
• AI Enthusiasts: Free use of local open-source models without API fees.

DevIO represents an important development direction for mobile AI apps—enjoying the capabilities of large language models without giving up control over data and privacy. Its multi-provider architecture and local-first design form a practical and flexible solution. As the hardware threshold for local models decreases, it will gain wider adoption among privacy-conscious users.