mini me desktop figurine — AI photo prompt example
Mini MeWorks withGeminiChatGPTDreamina

Mini Me Desktop Figurine AI Prompt

Upload one clear selfie first — this prompt keeps your own face.

PROMPT

Create an ultra-realistic 4:5 portrait of the person in the uploaded photo, preserving their exact facial features, skin tone and hairstyle. Reimagine them as a polished 3D miniature figurine standing on a walnut desk beside a keyboard, sketchbook and tiny plant, wearing their original outfit under warm window light. Shot like premium product photography on a 70mm macro lens at f/4, eye-level close composition with convincing scale and soft background blur. Natural skin texture, visible pores, film-like tones, no plastic smoothing, no text, no logos, no watermark.

Copied it but the result was off? Tell us which prompt and what went wrong.

How to use it

Copy the prompt above, upload one clear front-facing selfie, and paste both into Gemini or ChatGPT.

Works with

GeminiChatGPTDreamina
Back to all prompts

NOTES ON THIS PROMPT

This walnut-desk figurine starts from the uploaded face, then establishes its story through keyboard, sketchbook and tiny plant.

Why this prompt works

Light
warm window light crossing the walnut surface is the main structural decision for this frame. It separates the subject from keyboard, sketchbook and tiny plant without flattening the face, and it explains where every highlight and shadow should fall.
Lens
The specified 70mm macro lens at f/4 controls both proportion and context.
Style
premium collectible product photography gives the image its finish, while painted fabric, walnut grain and miniature skin detail keeps the result grounded.

Four one-line variations

Change one of these at a time. If you change several, you will not know which one moved the result.

Hair colour
For the desktop figurine, replace the hair instruction with 'a slightly tousled version of the uploaded hairstyle' and keep every face-identity phrase unchanged so only the styling moves.
Background
For the desktop figurine, swap the setting for 'a drafting table with clipped concept sketches' while retaining the same camera height, subject scale and wardrobe relationship.
Lens
For the desktop figurine, change the camera line to '100mm macro lens at f/5.6' for a different compression level without discarding the scene-defining objects.
Light
For the desktop figurine, rebuild the illumination as 'a cool monitor glow balanced by a small amber desk lamp' and preserve the original exposure on the face so the identity remains stable.

When the result goes wrong

Face warped
If the likeness drifts because of the small head and close desk objects, add 'symmetrical eyes, undistorted nose and jaw, exact facial proportions from the uploaded photo'. For the desktop figurine, put it directly after the identity sentence, before any scene description, so the model resolves the person before interpreting the environment.
Plastic skin
When surfaces become waxy, add 'visible pores, painted fabric, walnut grain and miniature skin detail, matte highlights, no beauty retouching'. For the desktop figurine, naming several real materials gives the renderer separate roughness cues and prevents one glossy finish from spreading across skin, clothing and props.
Broken hands
If the pose damages fingers, append 'both arms resting naturally beside the figurine body, natural knuckles, correct thumb placement, no fused digits'. For the desktop figurine, keep the hand action simple and physically supported by the nearby object; extra gestures compete with the already demanding face and setting.

How it compares to Mini Me Collectible on a Workbench

Compare this with mini me workbench collectible. This scene emphasizes a finished figure on a lived-in desk rather than a collectible being painted, and its 70mm macro lens at f/4 preserves that relationship. Choose this version when keyboard, sketchbook and tiny plant carries the narrative; use the comparison when its different environment better matches the mood you want to communicate.

See the other prompt

More prompts you might like

mini me held in a palm — AI photo prompt example
Mini MeWorks withGeminiChatGPT

Mini Me Held in a Palm AI Prompt

PROMPT

Create an ultra-realistic 4:5 portrait of the person in the uploaded photo, preserving their exact facial features, skin tone and hairstyle. Transform them into a tiny lifelike person standing confidently in an open human palm, wearing a casual layered outfit, with soft daylight wrapping around both figure and hand. Shot on a 90mm macro lens at f/3.2, centered full-body composition with realistic scale, detailed fingers and creamy room bokeh. Natural skin texture, visible pores, film-like tones, no plastic smoothing, no text, no logos, no watermark.

mini me in a miniature city street — AI photo prompt example
Mini MeWorks withGeminiDreaminaCapCut

Mini Me in a Miniature City Street AI Prompt

PROMPT

Create an ultra-realistic 4:5 portrait of the person in the uploaded photo, preserving their exact facial features, skin tone and hairstyle. Place them as a small 3D character walking through a handcrafted miniature city street, wearing modern streetwear among tiny storefronts, parked bicycles and glowing windows at blue hour. Shot on a 50mm tilt-shift lens at f/5.6, low eye-level composition with leading lines, practical lights and model-scale depth. Natural skin texture, visible pores, film-like tones, no plastic smoothing, no text, no logos, no watermark.

girl street documentary in black and white — AI photo prompt example
GirlWorks withGeminiChatGPT

Girl Street Documentary in Black and White AI Prompt

PROMPT

Create an ultra-realistic 4:5 portrait of the person in the uploaded photo, preserving their exact facial features, skin tone and hairstyle. Walking past a weathered fish market in understated streetwear, with wet concrete and crates behind them, under flat diffused overcast light. Shot on a 35mm lens at f/4, candid mid-stride vertical composition, black-and-white documentary style, rich silver tonality, deep blacks and 400 ISO grain. Natural skin texture, visible pores, film-like tones, no plastic smoothing, no text, no logos, no watermark.