Make a prompt directly from your images. Put in your image and get the approximate description/prompt from it to use in any text2image model to create similar images Advised to use it with CLIP and Diffusion guided models