Multiple Poses AI Photo Prompt (7 Poses, Same Outfit)
The multiple-poses photo is the collage filling Instagram stories: one person, seven poses, one frame — and it only works when every figure is the same person in the same clothes. This prompt locks face and outfit first, then names all seven poses, so the model cannot redesign you mid-row. Expression-strip and duo variations are inside.
One photo, seven versions of you: the viral multiple-poses collage with the same face, same outfit and same background in every figure — no clones drifting apart.
Use the uploaded photo as the ONLY face and outfit reference. Keep my face 100% identical in every figure — same eyes, nose, lips, skin tone, face shape and hairstyle — and keep my outfit exactly the same in every figure: same top, bottoms, shoes and colours, no added accessories. Create one wide photo showing the same person seven times in a single row against a plain light-grey studio wall, evenly spaced, full body visible in every figure. Each figure holds a different natural pose: 1) relaxed, facing the camera, small smile; 2) one hand on a hip, smiling; 3) arms crossed, looking off to the side; 4) hands in pockets, laughing down at the floor; 5) one hand on the chin, thinking, looking up; 6) hands behind the back, laughing out loud; 7) facing the camera with a surprised, playful expression. One soft studio light across the whole row, even exposure, natural skin texture, photorealistic, sharp focus on every figure. Negative: no different outfits, no different faces, no extra people, no text, no watermark, no beauty filter.
See what this prompt creates
It looks simple — seven of you, standing in a row — which is exactly why most attempts fall apart. The model treats each figure as a fresh person, so figure three gets a different jawline, figure five gets a different jumper, and the collage reads as seven strangers who happen to share a wall. The fix is not a better photo. It is a prompt that refuses to let anything change except the pose. It joins our viral photo trends shelf.
How to use it
- Upload one clear, full-body photo — front-facing, well lit, your whole outfit in frame. A headshot cannot tell the model what your shoes look like.
- Paste the prompt into ChatGPT or Gemini and generate. Do not trim the pose list: each numbered figure is a separate instruction.
- Check the row figure by figure at full size — faces first, then outfits. One drifted figure is easier to spot than to fix later.
- Want fewer figures? Delete whole pose sentences and renumber; never ask for "a few poses" and leave the count to the model.
Why this prompt works
Three decisions carry the whole image. The first is the double lock: face and outfit are locked in the same opening paragraph, before any pose is mentioned, so every figure inherits the same person and the same clothes. The second is naming every pose. "Seven different poses" is a wish; seven entries, one per figure, is a shot list the model can follow — and the poses differ in arms and gaze, not just smiles, so the row reads as seven moments rather than seven copies. The third is the single wall, single light rule. One background and one exposure is what makes seven figures read as one photograph instead of seven pasted portraits. The negative line guards the three classic failures: outfit changes, face drift, and a stranger wandering into the frame.
Two more ways to run it
- Expression strip: same pose every time, only the expression changes: neutral, smiling, laughing, thinking, surprised, winking, yawning. It works from a head-and-shoulders photo, which the full row does not.
- Duo version: upload two photos, name them in the lock, and ask for three figures each, alternating down the row, with the no-swap clause. Six figures is the limit for two identities.
Failure fixes: if one figure wears different clothes, the outfit lock was paraphrased — restore it word for word and regenerate. If faces drift, put your clearest photo first and add "every figure uses the uploaded face". If the row comes back as panels, add "one continuous photograph, no panels".
How to use this prompt
- Copy it — hit the copy button above to grab the raw prompt text.
- Fill the brackets — replace anything in [BRACKETS] with your own details.
- Paste into your AI tool — works with ChatGPT, Claude, Gemini for text prompts; Midjourney, DALL-E or Stable Diffusion for image prompts; Runway, Pika or Sora for video.
- Iterate — tweak one bracket at a time until the output is exactly right.
Frequently asked questions
How many poses can I ask for in one image?
Seven is the sweet spot. Beyond nine figures, faces shrink and drift — split a bigger set into two rows of five.
Does it work from a selfie or headshot?
Only for the expression strip. The full row needs a full-body photo — the model cannot invent your outfit below the frame.
One figure looks like a different person. What fixes it?
Regenerate first — variance fixes about half. If it persists, restate the face lock with one detail ('same freckles', 'same glasses').
ChatGPT or Gemini — which holds identity better here?
Both work and both drift occasionally. Gemini holds the face tighter; ChatGPT keeps the lighting more even. Try both, keep the better row.
