Summary

AI File Sorter is a cross-platform desktop application automating file organization via AI. It targets users needing efficient file management, offering local LLM support for offline, free categorization and a user-friendly interface for reviewing and confirming sorting actions.

How It Works

The application uses local (LLaMa, Mistral) or remote (ChatGPT 4o-mini) LLMs to categorize files based on names/extensions into user-defined categories. Its design prioritizes user control with a preview stage before execution, ensuring accuracy. Local LLMs enable offline operation and cost-free categorization.

Quick Start & Requirements

Installation requires compiling from source (git clone, make, sudo make install). Prerequisites include a C++ compiler, GTK+3/GTKMM3, and platform-specific build tools (MSYS2, Homebrew, apt/dnf/pacman). Optional CUDA Toolkit is recommended for GPU acceleration. An OpenAI API key is needed only for remote LLM usage.

Highlighted Details

• AI Categorization: Supports local (offline, free) and remote (ChatGPT) LLMs.
• User-Controlled Workflow: Preview dialog for reviewing/editing AI-assigned categories before sorting.
• Customizable Organization: Granular sorting with user-defined categories/subcategories.
• Cross-Platform: Windows, macOS, Linux support.
• Offline Capability: Full functionality with local LLMs.
• Secure API Key Handling: Encrypts API keys for remote LLMs.
• Update Notifications: Informs users of new releases.

Maintenance & Community

The README provides no details on specific contributors, community channels, or a public roadmap.

Licensing & Compatibility

Licensed under the GNU AFFERO GENERAL PUBLIC LICENSE (GNU AGPL). This strong copyleft license requires distributed derivative works or linked software to also be AGPL-licensed, potentially restricting proprietary integration.

Limitations & Caveats

The primary adoption barrier is the complex, multi-step source compilation process requiring significant technical expertise. The AGPL license may also limit suitability for certain commercial or closed-source integrations. No pre-compiled binaries are readily available.