ChatGPT Image Prompts with Examples

How to use these prompts

1. Choose an example with the framing you need. For a portrait, decide whether the result should be a close-up, a waist-up photograph or a full-body scene before adding styling.
2. In the prompt editor, describe the scene and identify what should stay fixed. If you use your own reference photograph, separate the person's appearance from the pose, clothing and background you want to change.
3. Copy the finished prompt into ChatGPT's image workflow. Attach any required reference there; copying a prompt does not attach your photographs. Describe one specific correction for the next iteration.

A simple prompt to adapt

Create an editorial portrait of one adult in a quiet bookshop. Waist-up framing at eye level, relaxed posture, soft window light from the left, natural skin texture. Keep the shelves softly out of focus and leave room above the head. Avoid text and logos.

When the result needs a revision

If the face changes too much, simplify the scene and clarify which reference supplies the identity. If a hand or object looks wrong, ask for a targeted correction instead of adding more style adjectives. Keep the original prompt so you can compare each revision.