🖼️ Sorterz — AI Image Organizer

Harness the power of Local Vision LLMs to intelligently categorize, rename, and organize your image library—100% offline and completely private.

Sorterz is a desktop application that uses open-source Vision Language Models (VLMs) to automatically understand the contents of your photos, extract EXIF data, and sort them into a clean, human-readable directory structure.

Instead of relying on cloud services that compromise your privacy and charge monthly fees, Sorterz runs entirely on your own hardware.

---

🔒 The Offline Advantage (Why Local AI?)

Organizing a lifetime of personal photos shouldn't mean uploading your private memories to a corporate cloud. Sorterz was built with a strict Offline-First philosophy:

• Absolute Privacy: Your images never leave your computer. Processing happens in your own GPU/CPU.
• Zero Recurring Costs: No API tokens to buy, no monthly cloud subscriptions. Sort 10 or 10,000 images for free.
• No Internet Required: Works beautifully whether you are on an airplane, off the grid, or just dealing with an internet outage.
• Total Control: You choose the model, the temperature, the limits, and the exact naming conventions.

✨ Key Features

• 🤖 AI Vision Processing: Generates context-aware descriptions for your images (e.g., "sunset.beach.friends", "red.sports.car").
• 📂 Smart Sorting & Renaming: Automatically renames files based on EXIF creation dates and AI descriptions (YYYY.MM.DD.three.word.desc.ext) and moves them into categorized subfolders.
• 🔌 Seamless Local Backends: Native integrations for LM Studio and Ollama.
  • Don't have them installed? Sorterz will auto-detect their absence and offer a 1-click install right from the UI.
• 💾 One-Click Model Management: Download and load vision models (like LLaVA, Qwen-VL, or Gemma) without touching a command line.
• 📅 Deep Metadata Extraction: Reads EXIF data (Original, Digitized) or falls back to filesystem creation/modification dates.
• 🕵️ Privacy Tools (EXIF Eraser): Optional feature to securely strip location and metadata from JPEGs before moving them.
• 🎮 Custom Prompts & Presets: Create your own prompting logic, or use built-in presets (like the Video Game / VRChat preset).
• ⚡ Live Telemetry: Watch the AI work in real-time with an integrated live image preview, activity log, and token processing speed (t/s) metrics.

---

🛠️ Supported AI Backends

Sorterz doesn't lock you into a single ecosystem. It supports:

1. LM Studio: Full CLI integration. Sorterz can start/stop the server, load models into VRAM, and unload them to free up memory.
2. Ollama: Directly pull tags and serve models via the Ollama local API.
3. Custom Endpoints: Connect to any external or custom OpenAI-compatible vision endpoint on your local network.

Recommended Vision Models

Sorterz provides built-in drop-downs for highly capable local VLMs with their approximate VRAM requirements:

• Gemma 4 E4B (~8 GB VRAM) — Great balance of speed and accuracy.
• Qwen 2.5-VL 7B (~6 GB VRAM) — Excellent at text recognition and precise details.
• LLaVA 13B (~9 GB VRAM) — A classic, highly capable vision model.
• ...or type in any custom HuggingFace / Ollama tag!

---

🚀 Getting Started

Prerequisites

• .NET 8.0 SDK
• A dedicated GPU (NVIDIA/AMD) or Apple Silicon is highly recommended for reasonable AI generation speeds.

Installation & Build

1. Clone the repository:

   git clone https://github.com/your-username/sorterz.git
   cd sorterz
   

2. Build and Run:

   dotnet build
   dotnet run --project Sorter
   

(Note: Sorterz is built using Avalonia UI and is fully cross-platform compatible across Windows, macOS, and Linux).

---

📖 How to Use Sorterz

1. Set Your Folders: Select the Source folder (where your messy images are) and an Output folder.
2. Choose a Backend: Select LM Studio or Ollama from the middle panel. If the indicator light is gray, click the button to install it.
3. Load a Model: Pick a model from the right panel and click Install Model / Start Server.
   • Tip: Enable "Limit Server" to automatically shut down any background models taking up VRAM.
4. Tweak Settings (Optional): Check "Erase EXIF" for privacy, or toggle "Custom Prompt" if you want the AI to format names differently.
5. Start Sorting! Press the big primary button at the bottom right. Sit back and watch the live preview and activity log as your library is perfectly organized.

---

🏗️ Technical Stack

• C# / .NET 8: Core application logic.
• Avalonia UI: Modern, cross-platform UI framework utilizing a dark industrial, terminal-inspired aesthetic.
• CommunityToolkit.Mvvm: Clean, efficient MVVM architecture.
• MetadataExtractor: Robust EXIF parsing across multiple file formats.
• Newtonsoft.Json: Payload serialization for local API communication.

---

🤝 Contributing

Contributions are welcome! If you want to add support for new backends, refine the UI, or improve the prompt logic:

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes (git commit -m 'Add amazing feature')
4. Push to the branch (git push origin feature/amazing-feature)
5. Open a Pull Request

📄 License

This project is open-source. (Please add your specific license here, e.g., MIT License).