Gerbil is a user-friendly tool that allows you to run Large Language Models (LLMs) on your local machine without the usual technical hassles. Built on KoboldCpp, which is derived from llama.cpp, it works across Windows, macOS, and Linux platforms. You can operate it entirely offline by importing pre-downloaded binaries. Gerbil supports both CPU systems and GPU acceleration through options like CUDA and Vulkan, giving flexibility depending on your hardware capabilities. Users can generate text and images with built-in presets for various workflows. The application integrates with HuggingFace, enabling users to browse models and download them directly. Advanced users can access over 70 command-line arguments through a modal interface, allowing for customization beyond what's available in the graphical interface. Running Gerbil in CLI mode is possible, although Windows users should install via the Setup.exe for optimal functionality. The app is privacy-focused, processing everything locally without sending data externally. Installation varies by operating system; Windows users can download a portable version or use the installer, while macOS users might need to adjust system settings to run the app due to Apple's security measures. For Linux users, particularly those on Arch, installation through the AUR simplifies updates. Gerbil's lightweight design keeps memory usage low, around 200MB of RAM and 100MB of VRAM for the GUI. In CLI mode, this drops significantly, making it suitable for users looking to minimize resource consumption while still leveraging advanced features.