Prompt Library

The Best Gemini AI Photo Prompts to Copy in 2026

28 copy-paste prompts

Upload a photo to Gemini, paste one of these prompts, and its image model (Nano Banana) does the rest. This is the master list — viral trends, professional headshots, couple edits, artistic styles, and photo repair, all tested to keep your real face intact.

In short: This page contains 28 copy-paste ready prompts, organized into 5 categories with a description and pro tip for each. The first 15 prompts are free instantly — no signup needed. Hand-curated and tested by the AI Academy team.

By Louis Corneloup · Founder, Techpresso
Last updated ·Hand-curated & tested by the AI Academy team

Prompts get you started. Tutorials level you up.

A growing library of 300+ hands-on AI tutorials. New tutorials added every week.

Start 7-Day Free Trial

Professional & LinkedIn Looks

6 prompts

Natural-Light Office Headshot

7/28

Using my uploaded selfie, create a professional headshot taken in a bright modern office: I'm wearing a navy crew-neck sweater over a collared shirt, standing near floor-to-ceiling windows with a softly blurred open-plan office behind me. Natural daylight from the side, relaxed confident half-smile, head-and-shoulders crop. Keep my exact facial features, age, and skin tone. Photorealistic, 1:1 for LinkedIn.

The approachable alternative to a formal studio headshot — daylight and a real office read warmer than a seamless backdrop.

💡

Pro tip: Match the outfit to your actual industry; 'navy sweater over a collared shirt' lands very differently than 'charcoal suit and tie'.

Founder Editorial Portrait

8/28

Transform my uploaded photo into a magazine-style founder portrait: seated on a simple stool against a dark grey textured wall, elbows on knees, hands loosely clasped, looking directly into the camera. Single dramatic softbox from the left, deep shadows on the right, slight film grain, muted color grade. Keep my exact facial features, age, and skin tone. 4:5 editorial portrait.

The moody profile-feature look used for press pages and conference bios — composed, serious, and clearly not a webcam crop.

💡

Pro tip: If the pose comes out stiff, add 'relaxed shoulders, weight slightly forward' — Gemini defaults to a bolt-upright posture.

Speaking on Stage

9/28

Using my uploaded photo, create an image of me speaking on a conference stage: holding a clicker mid-gesture, wearing a lanyard badge, lit by a warm stage spotlight against a dark blue background with a blurred slide screen glowing behind me. Shot from the audience at a slight low angle, shallow depth of field, photorealistic event photography. Keep my exact facial features, age, and skin tone.

Instant speaker-page credibility shot. The lanyard, clicker, and blurred slides are the details that make it read as a real event photo.

💡

Pro tip: Keep the slide screen 'blurred and unreadable' — asking for visible slide text is the fastest way to get garbled letters that expose the edit.

Creative Professional on Color

10/28

Turn my uploaded selfie into a studio portrait for a creative portfolio: smart-casual outfit in earth tones, seated backwards on a chair against a solid burnt-orange seamless backdrop, soft even lighting, genuine mid-laugh expression. Clean, modern, slightly bold — like an agency team page. Keep my exact facial features, age, and skin tone. 4:5.

For designers, writers, and marketers who want personality on their about page without losing polish.

💡

Pro tip: Swap the backdrop color to match your brand palette — Gemini holds solid seamless colors accurately when you name them precisely.

Match Our Team-Page Style

11/28

Here are two photos: the first is an existing headshot from my company's team page, the second is my selfie. Create a headshot of me from the second photo that exactly matches the style of the first: same backdrop color, same lighting direction and softness, same crop, same color grade. Keep my exact facial features, age, and skin tone — only copy the photographic style, not the person.

Solves the real problem: joining a company whose team photos were all shot in one session you missed.

💡

Pro tip: Label the images clearly — 'only copy the photographic style, not the person' prevents Gemini from blending the two faces.

Casual Photo to Interview-Ready

12/28

Keep my face, expression, and pose from the uploaded photo exactly the same, but change my hoodie to a pressed light-blue button-down shirt and replace the messy bedroom background with a clean, softly blurred neutral wall. Match the new shirt and backdrop to the existing lighting on my face so nothing looks pasted. Keep my exact facial features, age, and skin tone.

The minimum-effort fix: one decent casual photo becomes an application-ready portrait without changing who you are.

💡

Pro tip: Anchoring on 'match the existing lighting on my face' is what stops the swapped shirt from looking like a cardboard cutout.

Couple & Family Edits

5 prompts

Anniversary Dinner Portrait

13/28

Here are two photos: the first is me, the second is my partner. Merge us into one photorealistic image seated together at a candlelit restaurant table, leaning toward each other and laughing, wine glasses and warm bokeh lights in the background. Match the candlelight on both faces so it reads as one real photo. Keep both of our exact facial features, ages, and skin tones. 4:5.

Turns two ordinary selfies into a date-night portrait you never actually took — the candlelight matching is what makes the merge believable.

💡

Pro tip: If one face looks brighter than the other, ask Gemini to 'rebalance the candlelight so both faces share the same warmth' instead of regenerating.

Three Photos, One Family Portrait

14/28

Combine these three uploaded photos — me, my spouse, and our two kids — into a single relaxed family portrait on a living-room sofa with warm window light. Arrange us naturally: parents seated, kids leaning in, everyone mid-smile as if reacting to a joke. Match lighting, skin tones, and white balance across everyone. Keep every person's exact facial features, age, and skin tone from their reference photo.

The hardest merge Gemini does well: multiple sources, one believable scene. Ideal when getting everyone in one room is the actual problem.

💡

Pro tip: Describe the seating arrangement yourself — left to right if needed — or Gemini will invent one and sometimes duplicates a person.

Traditional Pre-Wedding Portrait

15/28

Using the two uploaded photos of me and my partner, create a traditional pre-wedding portrait: dress us in coordinated wedding attire from our culture (a deep red lehenga with gold embroidery and an ivory sherwani), posed in a palace courtyard at dusk with string lights and marigold garlands. Romantic, editorial, photorealistic. Keep both of our exact facial features, ages, and skin tones.

A full engagement-shoot look from two phone photos. Swap the attire description for any tradition — the structure carries over.

💡

Pro tip: Name the garments specifically (lehenga, sherwani, hanbok, barong) — generic 'traditional clothes' produces a confused costume mashup.

Matching-Sweater Family Card

16/28

Using my uploaded family photo, create a studio holiday-card portrait: everyone in matching cream knit sweaters against a warm beige seamless backdrop, soft even lighting, arranged close together and genuinely smiling. Leave clean space at the top for a greeting to be added later. Keep everyone's exact facial features, ages, and skin tones — do not add or remove anyone. 5:7 portrait.

The annual-card photo without wrangling everyone into a studio. The reserved headline space makes it drop straight into a card template.

💡

Pro tip: 'Do not add or remove anyone' matters — Gemini occasionally gifts large families a bonus child.

Couple Photo-Booth Strip

17/28

Using the two uploaded photos of me and my partner, create a vertical 4-frame photo-booth strip on a white border: frame 1 smiling at the camera, frame 2 making silly faces, frame 3 one kissing the other's cheek, frame 4 laughing mid-collapse. Slightly grainy flash photography look, consistent booth curtain background. Keep both of our exact facial features, ages, and skin tones in every frame.

Four consistent frames in one generation — a believable booth strip from two separate selfies, great for anniversaries and prints.

💡

Pro tip: Number each frame's action in the prompt; unnumbered lists make Gemini repeat the same pose four times.

Like these prompts? There are full tutorials behind them.

Learn the workflows, not just the prompts. 300+ easy-to-follow tutorials inside AI Academy — and growing every week.

Try AI Academy Free

Artistic Styles

6 prompts

Neo-Noir Film Still

18/28

Transform my uploaded photo into a cinematic neo-noir film still: standing on a rain-slicked city street at night, neon signs reflecting pink and cyan in the puddles, wet hair, trench coat, half my face lit by a flickering storefront. Anamorphic widescreen framing with subtle lens flare, heavy atmosphere, film grain. Keep my exact facial features, age, and skin tone. 21:9.

The moody movie-frame look — the rain reflections and split lighting do the cinematic heavy lifting.

💡

Pro tip: Request '21:9 widescreen with space beside me' — the negative space is what makes it feel like a film frame instead of a costume portrait.

High-Fashion Editorial Cover

19/28

Turn my uploaded selfie into a high-fashion magazine editorial: avant-garde structured blazer with exaggerated shoulders, slicked-back hair, intense direct gaze, shot against a stark white cyclorama with hard studio flash and a crisp shadow. Leave space at the top where a masthead would sit. Keep my exact facial features, age, and skin tone. 4:5, sharp and glossy.

The fashion-week treatment — hard flash and a fearless expression most people would never attempt in a real shoot.

💡

Pro tip: Hard flash is unforgiving; if it exaggerates skin texture, add 'clean editorial retouching, natural skin' to pull it back.

Classical Oil Painting

20/28

Repaint my uploaded portrait as a 17th-century oil painting: dark umber background, dramatic single-source candlelight from the left, rich glazed shadows, visible brushwork and craquelure, formal period clothing with a white collar. Present it inside an ornate gilded frame photographed on a museum wall. Keep my exact facial features, age, and skin tone beneath the painted style.

Old-master lighting plus the museum-wall framing — the gilded frame is what makes people stop scrolling.

💡

Pro tip: The phrase 'visible brushwork and craquelure' is what separates a real painted texture from a smooth photo filter.

Loose Watercolor Portrait

21/28

Convert my uploaded photo into a loose watercolor illustration: wet-on-wet washes in indigo and burnt sienna, deliberate white paper left showing, paint drips and blooms at the edges, facial features suggested with a few confident brushstrokes but still clearly recognizable as me. Cold-pressed paper texture, square 1:1.

The opposite of photorealism — an artist's sketchbook study where the unfinished edges are the charm.

💡

Pro tip: Say 'still clearly recognizable as me' — fully loose watercolor briefs can drift into a generic face.

Fine-Art Black & White

22/28

Transform my uploaded portrait into a fine-art black-and-white photograph: deep blacks, bright controlled highlights, a single beam of window light across my eyes, the rest of the frame falling into shadow. Strong texture in skin and fabric, no color anywhere, gallery-print quality. Keep my exact facial features, age, and skin tone rendered faithfully in monochrome.

A dramatic mono portrait built on one beam of light — the kind of frame photographers spend an afternoon rigging.

💡

Pro tip: Define where the light lands ('across my eyes') — vague 'dramatic lighting' gets you a generic grey vignette.

Hand-Drawn Anime Film Frame

23/28

Redraw my uploaded photo as a frame from a gentle hand-drawn anime film: soft painted background of a hillside town at golden hour, expressive simplified features that still resemble me, wind in the hair, warm pastel palette, subtle cel shading and painterly clouds. Keep my hairstyle, glasses, and outfit recognizable from the original photo. 16:9.

The cozy animated-film look — simplified but unmistakably you, thanks to the anchored hairstyle and outfit details.

💡

Pro tip: Anime styles compress facial detail, so anchor identity in hair, glasses, and clothing rather than micro-features.

Photo Repair & Enhancement

5 prompts

Colorize a Black & White Portrait

24/28

Colorize this black-and-white photo of my grandparents with realistic, period-accurate colors: natural skin tones, plausible 1950s clothing colors, and an environment colored to match the era. Do not alter their facial features, expressions, pose, or the composition in any way — add color only, and keep the original grain.

Pure colorization with identity locked down. Keeping the original grain is what stops it from looking like a modern reshoot.

💡

Pro tip: If you know real details — 'her cardigan was mint green' — state them; Gemini applies known colors faithfully instead of guessing.

Rescue a Blurry Photo

25/28

This uploaded photo is slightly blurry and underexposed. Sharpen the faces naturally, recover detail in the shadows, reduce noise, and correct the white balance — but do not change anyone's facial features, expressions, body shape, or position. The result should look like the same photo taken with a steadier hand and better light.

The one-prompt fix for the almost-great shot. The final sentence frames it as correction, not reinvention.

💡

Pro tip: Enhancement prompts are where Gemini quietly beautifies people — the explicit 'do not change facial features or body shape' clause is non-negotiable.

Fix a Backlit Silhouette

26/28

In my uploaded photo, the subject is backlit against a bright window and has become a dark silhouette. Brighten the subject so their face and clothing are clearly visible with natural skin tones, while keeping the bright window background from blowing out completely. Keep the person's exact facial features and the original composition — fix the exposure balance only.

Recovers the classic ruined shot: a face lost against a sunset or window. Reads like a skilled dodge-and-burn, not an edit.

💡

Pro tip: Ask to balance, not just brighten — 'brighten the photo' alone washes out the sky you were posing in front of.

Repair a Torn Printed Photo

27/28

Restore this scanned printed photo: repair the tear running through the corner, remove the creases, water stains, and dust spots, and rebuild the small missing area along the edge to match the surrounding scene. Do not modify any faces, clothing, or the composition — repair the physical damage only, preserving the photo's original colors and grain.

Physical-damage triage for shoebox photos: tears, creases, and stains gone, with the people untouched.

💡

Pro tip: Point out each damaged area by location ('the tear through the top-left corner') so the repair targets damage and never wanders onto a face.

Upscale a Tiny Photo for Print

28/28

This uploaded image is a small, low-resolution photo. Recreate it as a high-resolution version suitable for an 8x10 print: cleanly defined edges, natural skin texture instead of smudged pixels, and no compression artifacts. Keep every person's exact facial features, the colors, and the composition identical to the original — increase clarity and resolution only.

Turns a thumbnail-sized keepsake — an old profile picture, a cropped group shot — into something you can actually frame.

💡

Pro tip: Name the print size you want; 'suitable for an 8x10 print' gives Gemini a concrete sharpness target instead of a vague 'make it HD'.

Frequently Asked Questions

Open the Gemini app or gemini.google.com, tap the + or image icon to upload your photo, paste a prompt, and send. Gemini's image model (nicknamed Nano Banana) returns the edit in seconds. For multi-photo prompts — couple merges, style matching — upload all the photos in the same message and refer to them as 'the first photo' and 'the second photo'.
Nano Banana is the community nickname for the image generation and editing model built into Gemini. When you upload a photo to Gemini and ask for an edit, Nano Banana is what does the work — so 'Gemini AI photo prompts' and 'Nano Banana prompts' largely refer to the same thing. We keep a dedicated list of model-specific tricks like multi-image fusion in our Nano Banana prompts page.
Two common causes: the prompt never told Gemini to preserve your identity, or your source photo was low quality. Every transformation prompt should include 'keep my exact facial features, age, and skin tone' — without it, Gemini tends to beautify, slim, and de-age faces. Start from a sharp, well-lit, front-facing photo, and if a result drifts, reply with 'make the face match my uploaded photo exactly' instead of regenerating from scratch.
Yes — the free tier of Gemini includes photo upload and image editing, with daily generation limits. Paid plans (Google AI Pro and Ultra) raise the limits, process faster, and get priority access to the newest image models. For most people working through these prompts, the free tier is plenty to start.
Yes, and it is one of its strengths. It can merge separate photos of two or more people into one scene, edit group shots while preserving every face, and apply one photographic style across a whole family portrait. The key is to state how many people there are and add 'do not add or remove anyone' — group edits are where extra people occasionally appear.
Gemini declines edits involving minors in some contexts, public figures, and anything it reads as deceptive or explicit. Rephrase toward a clearly creative goal ('a painted portrait in 17th-century style') rather than an impersonation, make sure you are editing your own photo, and avoid prompts that change a person's apparent identity. Most refusals on this list trace back to a low-quality upload being misread, so a clearer photo often fixes it too.

Prompts are the starting line. Tutorials are the finish.

A growing library of 300+ hands-on tutorials on ChatGPT, Claude, Midjourney, and 50+ AI tools. New tutorials added every week.

7-day free trial. Cancel anytime.