TagGUI is a desktop application that helps creators manage and annotate image datasets for training generative AI models. It provides keyboard-friendly tagging, autocomplete from frequently used tags, and integrates with Stable Diffusion token counting. The application also supports automatic caption and tag generation using AI vision models including CogVLM, Florence-2, and LLaVA. Tags and captions are stored in text files alongside images, making datasets ready for training pipelines.