A text-to-image model is a machine learning model which takes an input natural language description and produces an image matching that description. From Wikipedia
