Black Friday SaleEnds in0 00:00:00
Reference Pose Remix
A ChatGPT image prompt written and tested by Visual AI Club.
Copy the full prompt with the button below and paste it into ChatGPT.
The guide further down explains each part and what to change.Prompt guide
Written as a structured JSON brief, this prompt borrows only the pose, camera angle and accessories from your reference and builds a new photo around a fictional woman in a different room. It is meant for reusing a composition you like, not for copying the person in it.
By the Visual AI Club team · How these guides are written
Edit the subject block before anything else
Set hair color, makeup and a clearly different face in the subject and appearance fields so the new woman is truly new.
Swap the room in the environment field
Try a sunlit concrete loft or a wood-paneled cabin, keeping the mirror mentioned if your reference is a mirror selfie.
Change one block per run
Big shifts such as a neon night scene can loosen the framing, so alter a single field and compare.
Do this
The new woman resembles the reference person
Write two or three clearly different features, such as hair color, face shape and age range, into the appearance field.
The phone gesture no longer makes sense
Put the mirror back into the environment description so the selfie pose has a reason.
The pose drifted after a lighting change
Return to the original lighting and change only the room or the clothing colors.
If you want
identity: a fictional adult man, with matching changes in appearanceIf you want
environment: minimal white studio with a full-length mirrorIf you want
environment: small cabin with warm wood panelingThe key line asks for a completely different fictional adult woman. Everything about the body position stays: crouch or sit, leg placement, torso angle, the hand holding the phone. Everything about identity changes. The important list repeats this several times so the model does not quietly reuse the original face.
This split is useful when you have a pose reference from your own shoot or a stock image you are licensed to use, and you want a fresh variation for a mood board or a fashion layout.
reference limits what is copied: pose, angle, composition, accessories, overall style.subject sets the new woman's look; edit hair color or makeup here.environment swaps the setting, with a hotel room or modern apartment as the example.clothing keeps the silhouette but changes details and colors.camera locks in a vertical smartphone mirror selfie with natural phone image quality.The environment block is the easiest place to experiment. Try a sunlit loft with concrete floors, a cozy cabin with wood paneling, or a minimal white studio. Keep the mention of a mirror if the reference is a mirror selfie, otherwise the phone gesture stops making sense.
Big lighting changes, such as moving to a neon-lit night scene, can pull the model away from the original framing. Change one thing per run so you can see what caused a shift.
Only upload pictures you took, pictures of people who agreed, or images you have rights to. The prompt is built to avoid copying someone's identity, and it works best if you keep it that way rather than steering the new face back toward the original.
Avoid presenting the output as a real photo of a real person; it is a staged, generated image.
Yes. Edit the identity and appearance fields to describe a fictional adult man. Accessories like rings and shoes will adapt as well.
Strong reference photos can bleed through. Describe two or three clearly different features, such as hair color, face shape and age range, in the appearance field.
Next prompt · ChatGPTGTA VI Artwork PromptOpenLearn AI Academy
Copy prompts that work, then learn why they work. The Learn AI Academy teaches AI image and AI video creation step by step: writing prompts, light and camera, consistent characters, product shots and directing a short film with Seedance.
Open the Learn AI Academy