Show HN: AI File Sorter: Organize Files and Folders with AI (Local LLMs)

AI File Sorter is a powerful, cross-platform desktop application that automates file organization. Featuring AI integration and a user-friendly GTK-based interface, it categorizes and sorts files and folders based on their names and extensions. The app intelligently assigns categories and, optionally, subcategories, which you can review and edit before confirming. Once approved, the necessary folders are created, and your files are sorted accordingly. The app uses local (LLaMa, Mistral) and remote (ChatGPT 4o-mini) LLMs for this task, depending on your choice.

• Features
• Requirements
• Installation
  • Windows
  • MacOS (Apple Silicon)
  • Linux
• API Key, Obfuscation, and Encryption
• Uninstallation
• How to Use
• Sorting a Remote Directory (e.g., NAS)
• Contributing
• License
• Credits
• Donation

• AI-Powered Categorization: Classify files intelligently using either a local LLM (LLaMa, Mistral) or a remote LLM (ChatGPT), depending on your preference.
• Offline-Friendly: Use a local LLM to categorize files entirely e—no internet or API key required. Customizable Sorting Rules: Automatically assign categories and subcategories for granular organization.
• Intuitive Interface: Lightweight and user-friendly for fast, efficient use.
• Cross-Platform Compatibility: Works on Windows, macOS, and Linux.
• Local Database Caching: Speeds up repeated categorization and minimizes remote LLM usage costs.
• Sorting Preview: See how files will be organized before confirming changes.
• Secure API Key Encryption: When using the remote model, your API key is stored securely with encryption.
• Update Notifications: Get notified about updates—with optional or required update flows.

• Operating System: Windows, macOS, or Linux with a stable internet connection.

• C++ Compiler: A recent g++ version (used in Makefile).

• Platform specific requirements:

  • Windows: MSYS2 / MINGW64 with some requirements.
  • MacOS: brew to install some requirements. Xcode.

  Optional:

  • Git: For cloning this repository. You can alternatively download the repo in a zip archive.
  • OpenAI API Key: Not needed for local LLMs, but needed to download the models.

File categorization with local LLMs is completely free of charge.

If you want file categorization with ChatGPT, you will need to get an OpenAI API key and add a minimal balance to it for this program to work. Categorization is quite cheap, so $0.01 will be enough to categorize a relatively large number of files. The instructions on how to integrate your API key into the app are given below. You can also download a Release version, which has an embedded API key.

First, you need git. Download git-scm for Windows from git-scm.com and install it. Verify the installation in cmd or powershell with

You can also now launch Git Bash from Start Menu.

Important: Install all the packages specified below in their default installation directories. Otherwise, you'll need to adjust the paths in build_llama_windows.ps1 and Makefile.

1. Install MSYS2. Make sure to launch as Administrator MSYS2 MINGW64, NOT MSYS2 MSYS. If you don't launch it as Administrator, make install won't work. You may run it as regular user if you don't need to use make install in the last step.
2. Update MSYS2 packages: pacman -Syu.
3. Install dependencies:

4. Install Visual Studio Community. Imporant: Be sure to check the Dekstop development with C++ workflow, under which the Windows 1x SDK (depending on your Windows version) should be checked.

5. Optional: If you want to compile AI File Sorter with CUDA (for Nvidia GPUs only) then download and install CUDA Toolkit. CUDA significantly increases the compute speed of local LLMs. Otherwise, AI File Sorter will fall back on either OpenCL or OpenBLAS (CPU only, but faster than without it).

6. Install Python.

7. Launch Developer PowerShell for VS 2022 and cd path\to\cloned\github\repo\app\scripts (e.g., cd C:\Users\username\repos\ai-file-sorter\app\scripts). Check that the script build_llama.windows.ps1 contains the paths in accordance with your versions of CUDA and Visual Studio tools. In particular, check the version numbers in these lines:

   $cudaRoot = "C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v12.9" -DCMAKE_C_COMPILER="C:/Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/14.44.35207/bin/Hostx64/x64/cl.exe" ` -DCMAKE_CXX_COMPILER="C:/Program Files/Microsoft Visual Studio/2022/Community/VC/Tools/MSVC/14.44.35207/bin/Hostx64/x64/cl.exe"

   Save any changes.

8. In Developer PowerShell for VS 2022, run

   If you have CUDA:

   powershell -ExecutionPolicy Bypass -File .\build_llama_windows.ps1 cuda=on

   If you don't have CUDA:

   powershell -ExecutionPolicy Bypass -File .\build_llama_windows.ps1 cuda=off

9. Optional (not needed if you want to use only local LLMs for file sorting). Go to API Key, Obfuscation, and Encryption and complete all steps there before proceeding to step 6 here. The app won't work otherwise.

10. Go back the MSYS2 MINGW64 shell (ensure you ran it as Administrator, otherwise make install won't work). Go to the cloned repo's app/resources path (e.g., /c/Users/username/repos/ai-file-sorter/app/resources) and run bash compile-resources.sh. Go to the app directory (cd ..).

11. Run make, make install and make clean. The executable AiFileSorter.exe will be located in C:\Program Files\AiFileSorter. You can add the directory to %PATH%.

1. Install Xcode tools: xcode-select --install.
2. Install Homebrew: /bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)".
3. Add these lines to your ~/.zshrc file:

Then source ~/.zshrc or source ~/.bashrc.

4. Install dependencies:

   brew install gcc [email protected] cairo at-spi2-core [email protected] gtk+3 gtkmm3 [email protected] [email protected] pango harfbuzz glib gettext curl jsoncpp sqlite3 openssl@3 pkg-config libffi expat xproto xorgproto fmt spdlog adwaita-icon-theme hicolor-icon-theme brew install --cask font-0xproto

5. Go to API Key, Obfuscation, and Encryption and complete all steps there before proceeding to step 6 here. The app won't work otherwise.

6. Go to app/resources and run ./compile-resources.sh. Go back to the app directory.

7. Run make, sudo make install, make clean. Then you can launch the app with the command aifilesorter.

2. Go to API Key, Obfuscation, and Encryption and complete all steps there before proceeding to step 6 here. The app won't work otherwise.

3. Go to app/resources and run ./compile-resources.sh. Go back to the app directory.

4. Run make, sudo make install, make clean. Then you can launch the app with the command aifilesorter.

Before compiling the app:

1. Get an OpenAI API key from the OpenAI website.
   A minimal balance is required in your OpenAI API account for the app to function.

2. Generate a 32-character random secret key, e.g., using this tool.

   Important: If you're compiling on Windows, make sure there is NO = in the generated key! If one or more = are there, regenerate the key! Important: If you're compiling on Windows, it's probably best to avoid symbols due to possible unpredictable parsing issues.

   Your secret key could look something like sVPV2fWoRg5q62AuCGVQ4p0NbHIU5DEv or du)]--Wg#+Au89Ro6eRMJc"]qx~owL_X.

3. Navigate to the api-key-encryption folder, then make a file named encryption.ini with the following content:

   LLM_API_KEY=sk-... SECRET_KEY=your-generated-32-byte-secret-key

4. Run the compile.sh script in the same directory to generate the executable obfuscate_encrypt. due

5. Execute obfuscate_encrypt to generate:

   • Obfuscated Key part 1
   • Obfuscated Key part 2
   • Encrypted data (hex)

6. Update the application files:

   • Update app/include/CryptoManager.hpp with Obfuscated Key part 1:
     static constexpr char embedded_pc[] = "insert-obfuscated-Key-part-1-here";
   • Add the values to app/resources/.env as shown:
     ENV_PC=obfuscated-key-part2-value ENV_RR=encrypted-data-hex-value

7. Continue with Installation

In the directory app, run sudo make uninstall.

1. Launch the application (see the last step in Installation according your OS).
2. Select a directory to analyze.
3. Tick off the checkboxes on the main window according to your preferences.
4. Click the "Analyze" button. The app will scan each file and/or directory based on your selected options.
5. A review dialog will appear. Verify the assigned categories (and subcategories, if enabled in step 3).
6. Click "Confirm & Sort!" to move the files, or "Continue Later" to postpone. You can always resume where you left off since categorization results are saved.

Follow the steps in How to Use, but modify step 2 as follows:

• Windows: Assign a drive letter (e.g., Z: or X:) to your network share (instructions here).
• Linux & macOS: Mount the network share to a local folder using a command like:
  sudo mount -t cifs //192.168.1.100/shared_folder /mnt/nas -o username=myuser,password=mypass,uid=$(id -u),gid=$(id -g)

(Replace 192.168.1.100/shared_folder with your actual network location path and adjust options as needed.)

• Fork the repository and submit pull requests.
• Report issues or suggest features on the GitHub issue tracker.
• Follow the existing code style and documentation format.

• Curl: https://github.com/curl/curl
• Gtk+3: https://gitlab.gnome.org/GNOME/gtk
• Dotenv: https://github.com/motdotla/dotenv
• git-scm: https://git-scm.com
• Hugging Face: https://huggingface.co
• JSONCPP: https://github.com/open-source-parsers/jsoncpp
• LLaMa: https://www.llama.com
• llama.cpp https://github.com/ggml-org/llama.cpp
• Minstral AI: https://mistral.ai
• MSYS2: https://www.msys2.org
• OpenAI: https://platform.openai.com/docs/overview
• OpenSSL: https://github.com/openssl/openssl
• SoftPedia: https://www.softpedia.com/get/File-managers/AI-File-Sorter.shtml
• spdlog: https://github.com/gabime/spdlog

This project is licensed under the GNU AFFERO GENERAL PUBLIC LICENSE (GNU AGPL). See the LICENSE file for details.

Support the development of AI File Sorter and its future features. Every contribution counts!

• Donate via PayPal
• Bitcoin: 12H8VvRG9PGyHoBzbYxVGcu8PaLL6pc3NM
• Ethereum: 0x09c6918160e2AA2b57BfD40BCF2A4BD61B38B2F9
• Tron: TGPr8b5RxC5JEaZXkzeGVxq7hExEAi7Yaj

USDT is also accepted in Ethereum and Tron chains.