Camille Dubois

Mini-Me: Add a Tiny 3D Version of Yourself to Your Real Desk

Place a tiny, recognizable 3D version of yourself into a real photo of your desk or room, with correct scale, contact shadows and matched lighting.

At a glance

Best for: People making a fun 'mini me' post or team photo from their own workspace, who want the figure to look placed in the scene, not pasted on top.
You fill in: placeheightoutfitactionsurface

image-generation

photo-editing

social-media

Prompt

I'm uploading two photos. Photo 1 is me. Photo 2 is {{place}}.

Create a realistic version of Photo 2, unchanged, with a tiny version of me (about {{height}} tall) added to it, as if a miniature figure of me were really there.

The mini me:

• Clearly recognizable from Photo 1: same face, skin tone and hairstyle, plus glasses or facial hair if I have them. Wearing {{outfit}}.

• Real human proportions (not a chibi or bobblehead), with a slightly polished collectible-figure finish.

• {{action}}, standing on {{surface}}.

Make it believable:

• Scale it against the real objects nearby, such as a mug or keyboard, so the size reads instantly.

• Match the light direction, color temperature and softness of Photo 2.

• Add a soft contact shadow where the feet touch the surface.

• Match the camera's depth of field and grain.

Keep everything else in Photo 2 exactly as it is: no new objects, no tidying, no color changes. Keep Photo 2's aspect ratio.

Fill in

{{place}}: What the second photo shows.e.g. my desk at home
{{height}}: Figure height, given relative to something real is best.e.g. 10 cm, about the height of the coffee mug
{{outfit}}: What the mini you wears.e.g. the same gray sweater as in Photo 1
{{action}}: A simple pose or action involving an object in the scene.e.g. Leaning on the coffee mug and waving
{{surface}}: Exactly where it stands in Photo 2.e.g. the desk to the right of the keyboard

Example input

Upload 1: a clear, front-facing photo of me from the waist up in a gray sweater, evenly lit.
Upload 2: my home desk photographed at desk height in daylight: laptop, keyboard, a white coffee mug, a small plant, window light from the left.

place: my desk at home
height: 10 cm, about the height of the coffee mug
outfit: the same gray sweater as in Photo 1, with jeans
action: Leaning on the coffee mug and waving at the camera
surface: the desk to the right of the keyboard

Written for Gemini (Nano Banana), which accepts several images in one prompt. It also works in ChatGPT Images 2.0.

2710

Why it works

This trend lives or dies on compositing realism, so most of the prompt is about making the figure belong in the scene. Giving a size relative to a real object ("about the height of the coffee mug") is far more reliable than centimeters alone, and naming nearby objects as scale references lets the viewer read the size instantly.

The contact shadow, matched light direction and color temperature, and matched depth of field are the three cues the eye uses to spot a paste-in. "Real human proportions" avoids the default big-head chibi style. The final line, keep everything else exactly, stops the model from quietly tidying or restyling your room, which is common when a second image is involved.

When not to use it

A background photo taken from far away or at a steep angle makes scale hard to get right, so shoot the surface close to its height. Small figures lose facial detail, and likeness drift is common below about 5 cm. The background photo can reveal your home, screens or documents, so check it for anything private before uploading. Don't put other people into scenes without their consent.

Comments (0)

No comments yet. Be the first to share your thoughts!