Pose Prompts for AI Image Generators

Copy a pose prompt that describes the body position clearly, then open the pose page for Midjourney and Stable Diffusion variants, a negative prompt and an OpenPose skeleton for ControlNet.

How to write a pose prompt that works

AI image models follow camera framing and body position better when you describe them literally: say which leg is forward, where the hands are and what the torso is doing, and name the shot (full body, three-quarter view) up front. Short, concrete phrases beat long lists of adjectives.

Text alone rarely locks a pose. For an exact match, use the OpenPose skeleton from the pose page with a ControlNet OpenPose model, keep the prompt for style and subject, and add the negative prompt to cut extra limbs and broken hands.