OmniParser, turn your LLM into GUI agent
Try on garments on virtual models
Generate text based on your input
Generate images from text descriptions