Image Generation Guides

• Common to models - These models can create people, animals, machines, and landscapes. In case of people, upper body images come out well, but when generating full-body images from a distance, facial details may not be rendered accurately.
• Image models “A, B, C, and E” are suitable for depicting attractive men and women. Unless a specific race is specified, most are depicted as Westerners.
• The image model “E” is best suited for expressing English letters in image and animals, machines, various range of realistic objects. However, there are difficulties in reproducing complex human movements. And it takes a little longer to generate than other models.
• The image model “B,” due to its learned characteristics, it is not suitable for expressing animals, and people must appear in the image.
• Image models “K, L, M, N, and O” generate animated images. Depending on the user's input, non-animated characters may also appear. Each model has a different style, so please test them to see which one you like best. These models are subject to change at any time as we search for better models.