Improved interpretation of AI prompts

In the case of robots, monsters, and fighting characters
I want to make a picture of a confrontation.
However, when multiple characters appear, they usually get mixed up and make a strange picture.
Is there any way to improve the AI's interpretation and understanding of these prompts?
I get bored when the number of pictures I can create is limited. If you don't want to be limited to beautiful girls and erotic images,
If you don't want to be limited to beautiful girls and erotic images, please make it easier to create a variety of pictures.