Your photo on the left, the edit on the right

AI Photo Editing Prompts

Upload a photo, paste one prompt, change one thing. These AI photo prompts are sorted by what you want to do with a picture you already have, and each one sits next to the photo it started from. Try any of them free in Nano Banana or Flux AI.

25 edits

each beside its photo

13 kinds of edit

background to age

3 editing models

one prompt, one change

One photo, three different edits

Every row starts with the photo that was uploaded. Each edit after it was made from that same photo with its own prompt, so you can see how far one picture goes.

  1. Step 1 · Your photo
    A smiling grey-haired grandmother in a flowered apron and beige cardigan folds dumplings at her kitchen counter, a steaming pot and rice cooker behind her

    the photo you start from: one face, morning light

    A candid phone photo of a Korean grandmother in her seventies standing at her small home kitchen counter, folding dumplings on a floured wooden board. Grey hair pulled back in a low bun, reading glasses pushed up on her head, a flowered apron over a beige cardigan. She looks up at the camera with a warm, crinkly smile. Morning light from a window on the left, steam rising from a pot on the stove, jars of spices and a rice cooker behind her. Shot at eye level on a phone, natural colors, slight softness, framed from the waist up with space above her head. Exactly one person. No text, no watermark.

  2. Step 2 · Edit
    The same grandmother in her kitchen, now in a teal dress with a white lace collar and a small pendant

    a new dress, the kitchen and the smile kept

    Change the older woman's flowered apron and beige cardigan to an elegant teal linen dress with a white lace collar and a thin gold brooch. Keep the kitchen, the dumplings on the floured board, her grey bun, the glasses on her head and the morning light unchanged, while keeping the same facial features and warm smile.

  3. Step 3 · Edit
    The grandmother folding dumplings painted as a loose watercolor with soft washes, the kitchen shelves and rice cooker kept

    the same moment as a watercolor print

    Change the style of this photo to a loose watercolor painting on textured cold-press paper: soft washes of color bleeding at the edges, fine pencil lines showing through, white paper left for highlights, a gentle warm palette. Keep the older woman with the grey bun folding dumplings, the kitchen and the composition where they are, while keeping the same facial features and warm smile.

  4. Step 4 · Edit
    The grandmother in her kitchen shown decades younger, dark hair tied back, the same apron and smile

    the same woman, decades younger

    Change the older woman to how she would have looked at about twenty-five: smooth skin, glossy black hair in the same low bun, no reading glasses, the same warm crinkly smile. Keep the flowered apron, the beige cardigan, the dumplings on the floured board, the kitchen and the morning light unchanged, while keeping the same facial features, eye shape and expression.

Change the background

Good for a profile picture taken somewhere dull, or a headshot without a studio. Tip: ask for the new background's blur and light to match the old one, or the person looks pasted on.

A teenage girl with a long black ponytail and a white tee sits on a skatepark ledge with a skateboard on her lap, palm trees and the sea behind her

Girl with a skateboard by the sea, the photo

Nano Banana

A phone photo of a Mexican teenage girl about seventeen sitting on a low concrete ledge at a beachside skate spot, her skateboard across her knees. Long black hair in a high ponytail, a loose white tee, baggy faded jeans, scuffed canvas sneakers, a small plaster on one elbow. She squints into the late afternoon sun with a relaxed grin. Palm trees, a pale concrete wall and a strip of ocean behind her. Warm natural light, slight lens flare, framed from the knees up with space above her head. Exactly one person, one skateboard. No logos, no text, no watermark.

The same girl with a skateboard on her lap, now on the rim of a graffiti skate bowl under a pink sunset sky

New background: skatepark at sunset

Flux Kontext

Replace the beach and palm trees behind the girl with the black ponytail with an empty concrete skatepark bowl at sunset, colorful graffiti on the far walls and a pink and orange sky. Keep her sitting on the same ledge with the skateboard across her knees, keep her clothes, pose and the warm side light, while keeping the same facial features and grin.

The same bearded man in a navy jacket with a coffee cup, now standing in a bright office lobby with wood panels and a plant blurred behind

New background: bright office lobby

Flux Kontext

Replace the bus shelter and the rainy street behind the man in the navy suit with a bright modern office lobby: pale wood panels, a large window with soft daylight and a few green plants, slightly blurred. Change the light on him to soft even daylight that matches the new room. Keep his pose, suit, coffee cup and bag unchanged, while keeping the same facial features, beard and expression.

Restore and colorize old photos

Good for bringing a family print back: creases, tears, fading and black and white. Tip: put the steps in order (repair, then color, then sharpen) and ask to keep the film grain. Made with Nano Banana.

A faded, creased 1970s snapshot of two laughing women in a red diner booth, one in a floral blouse, one with an auburn afro and a mustard turtleneck

Damaged 1970s snapshot, the photo

Nano Banana

A scanned 1970s color snapshot, faded with age: colors shifted toward orange and teal, washed-out contrast, a torn upper corner, a white crease across one side, small scratches and dust specks, soft focus from an old instamatic camera. Two women friends in their late twenties sit side by side in a red vinyl booth at a roadside diner, laughing, one with feathered brown hair and a wide-collar floral blouse, the other with a curly auburn afro and a mustard turtleneck, two milkshakes with straws and a plate of fries on the formica table. Warm window daylight, a chrome napkin holder and a jukebox selector on the wall behind them. Exactly two people. No readable text, no watermark.

The two friends in the diner booth restored: crease, tear and scratches gone, colors warm and natural again

Restore a damaged 1970s snapshot

Nano Banana

Restore this damaged old photo of two friends so it looks like a freshly printed photograph. Remove the white crease line completely, along with every scratch, dust speck and the torn corner, rebuilding the hidden parts of the booth, the table and their clothes. Remove the faded orange and teal cast entirely: restore full contrast, deep blacks, clean whites and natural, vivid color, with true skin tones, the bright red vinyl booth, the floral blouse and the mustard turtleneck. Keep both women exactly as they are: the same faces, features, laughter, hair and poses, the same milkshakes, fries and composition. Keep a fine film grain. No crease, no scratches, no color cast. No text, no watermark.

A worn black and white studio portrait of a young man in a tweed flat cap, a suit and a tie, smiling slightly

1940s black and white portrait, the photo

Nano Banana

A scanned 1940s black and white studio portrait print of a young man in his early twenties, head and shoulders, wearing a tweed flat cap, a wool jacket and a knitted tie, looking slightly off camera with a shy smile. Soft studio light from the left, a plain mottled backdrop. The print is old: silvering at the edges, a torn lower corner, a few white dust specks, a faint crease across the jacket, slight blur. Black and white only. Exactly one person, the whole head and cap inside the frame. No text, no watermark.

The young man in the flat cap restored and colorized, with blue eyes, a grey tweed suit and a dark red knitted tie

Restore and colorize a 1940s portrait

Nano Banana

Restore and colorize this old black and white portrait. First repair it: fill in the torn lower corner, remove the crease, the dust specks and the silvering at the edges, and gently sharpen the face. Then add natural color: warm realistic skin, brown hair under a grey-brown tweed cap, a dark green wool jacket, a cream shirt and a burgundy knitted tie, a soft warm grey backdrop. Keep his face, shy smile, pose and the studio lighting exactly as they are, with fine film grain so it still looks like a real photograph from the 1940s. No text, no watermark.

Try on a new outfit

Good for seeing a look on yourself before you buy it. Tip: name the fabric and the color, such as teal linen with a lace collar, not something nicer. These run in Flux AI.

A smiling grey-haired grandmother in a flowered apron and beige cardigan folds dumplings at her kitchen counter, a steaming pot and rice cooker behind her

Grandmother in her kitchen, the photo

Nano Banana

A candid phone photo of a Korean grandmother in her seventies standing at her small home kitchen counter, folding dumplings on a floured wooden board. Grey hair pulled back in a low bun, reading glasses pushed up on her head, a flowered apron over a beige cardigan. She looks up at the camera with a warm, crinkly smile. Morning light from a window on the left, steam rising from a pot on the stove, jars of spices and a rice cooker behind her. Shot at eye level on a phone, natural colors, slight softness, framed from the waist up with space above her head. Exactly one person. No text, no watermark.

The same grandmother in her kitchen, now in a teal dress with a white lace collar and a small pendant

New outfit: teal dress with a lace collar

Flux Kontext

Change the older woman's flowered apron and beige cardigan to an elegant teal linen dress with a white lace collar and a thin gold brooch. Keep the kitchen, the dumplings on the floured board, her grey bun, the glasses on her head and the morning light unchanged, while keeping the same facial features and warm smile.

The same man at the rainy bus stop, his suit jacket and shirt swapped for a crisp white linen shirt

New outfit: white linen shirt

Flux Kontext

Replace the man's navy suit jacket and white shirt with a light off-white linen shirt with rolled sleeves and sand-colored chinos. Keep the bus stop, the coffee cup, the bag, his pose and the framing the same, while keeping the same facial features and beard.

Pet portraits from a phone photo

Good for the most shared edit of a pet: an oil painting or a school picture day. Tip: describe the animal's own marks, like the flipped ear or the white paws, so it stays your pet. Made with Nano Banana.

A scruffy brown and white dog with one ear flipped rests its chin on a cushion on a green velvet sofa beside a tennis ball and a remote

Scruffy dog on the sofa, the photo

Nano Banana

A casual phone photo of a scruffy medium-sized mixed-breed dog with a brown and white wiry coat, lying on a worn olive-green velvet sofa, chin resting on a cushion, looking at the camera with one ear flipped inside out. A crumpled knitted throw, a chewed tennis ball and a TV remote beside him. Soft afternoon light from a window, a living room slightly out of focus behind. Shot from a little above, natural colors, real fur detail. Exactly one dog. No people, no text, no watermark.

The scruffy dog painted as a grand old master oil portrait in a gilded frame, wearing a lace ruff and a velvet cape on a red cushion

Pet portrait: 17th century oil painting

Nano Banana

Paint this dog as the subject of a grand 17th century oil portrait. The same scruffy brown and white dog, with his wiry coat, his face and his one flipped ear, sits upright on a deep red velvet cushion wearing a small ermine-trimmed royal cloak and a lace collar, looking at the viewer with great dignity. Dark painted background, a heavy drapery on one side, dramatic warm light from the upper left. Visible brushstrokes, fine craquelure in the varnish, and an ornate carved gilt frame around the painting, hanging on a dark green wall. Exactly one dog. No text, no watermark.

A black and white tuxedo cat with green eyes sits upright on a windowsill between pots of basil, a brick street blurred behind the glass

Tuxedo cat on the windowsill, the photo

Nano Banana

A phone photo of a black and white tuxedo cat sitting upright on a sunny windowsill between potted herbs, white chest and white paws, green eyes looking straight into the lens with a serious expression. Late morning sun from the side, a bit of street and a brick building blurred through the window glass, a few basil leaves and a small watering can beside her. Natural colors, real fur texture, framed with space above the ears. Exactly one cat. No people, no text, no watermark.

The tuxedo cat posed for a 1990s school picture in an argyle sweater vest and a blue bow tie against a mottled blue backdrop with laser beams

Pet portrait: 1990s school picture day

Nano Banana

Make a 1990s school picture day portrait of this cat. The same black and white tuxedo cat with green eyes and the same serious expression sits on a little stool, head and shoulders, wearing a tiny knitted sweater vest over a collared shirt, framed like a school yearbook photo. Behind her the classic mottled blue and purple studio backdrop with a hint of laser light streaks. Soft flash from the front, slightly soft focus, warm color print look with fine grain. Exactly one cat, ears fully in the frame. No text, no watermark.

Remove people and things

Good for the stranger in the back of a holiday photo or the receipt on the table. Tip: name each thing to remove and what should fill the space it leaves. Try it in Nano Banana.

A laughing couple walks arm in arm along an autumn train platform, a blue train beside them and a stranger in a grey hoodie with a suitcase behind

Couple on a station platform, the photo

Nano Banana

A phone photo taken by a friend of a couple in their early thirties standing close together on a train station platform on an overcast autumn afternoon. A Black woman with dark curly shoulder-length hair in a camel coat and a South Asian man with short black hair and a navy rain jacket, both laughing, her arm through his, a small travel bag at their feet. Behind them a stopped commuter train, a platform canopy and, a little to the right, a stranger in a grey hoodie walking past with a rolling suitcase. Flat cool daylight, natural colors, framed from the knees up. Exactly three people. Signs blank, no readable text, no watermark.

The same couple on the platform with the stranger in the grey hoodie removed, the empty platform filled in behind them

Remove a stranger from the background

Flux Kontext

Remove the stranger in the grey hoodie with the rolling suitcase from the background and fill the space with the platform and the train behind, continuing the same lines and light. Keep the couple, their coats, the bag at their feet and everything else unchanged, while keeping the same facial features of the woman with curly hair and the man with short black hair.

A bowl of shoyu ramen with chashu, egg and nori on a wooden cafe table next to a phone, a receipt and chopsticks

Ramen on a cafe table, the photo

Nano Banana

A phone photo of a bowl of shoyu ramen on a small wooden cafe table by a window, shot from a 45 degree angle. Glossy brown broth, wavy noodles, two slices of chashu pork, a halved soft egg, nori, scallions and a spoon resting in the bowl, chopsticks on a ceramic rest. Also on the table: a smartphone face down, a crumpled paper receipt, a used napkin and a half-empty glass of water. Bright midday daylight from the window, a blurry street outside. Natural colors, real steam. Exactly one bowl. No readable text, no watermark.

The same bowl of ramen on a clean wooden table, the phone and receipt gone, only chopsticks on their rest beside it

Remove clutter from the table

Nano Banana

Clean up the table in this photo: remove the smartphone, the crumpled receipt, the used napkin and the glass of water, and fill those spots with the same wooden tabletop and grain. Keep the ramen bowl with all its toppings, the steam, the chopsticks on their rest, the window light and the camera angle exactly as they are. The result should look like the photo was simply taken before the table got messy.

Change hair color and style

Good for testing a dye or a cut on your own face first. Tip: when only the color should change, say the length and the parting stay as they are. The single edit is Flux AI; the grid of four cuts is Nano Banana. To try six cuts on your own selfie at once, use the AI hairstyle changer.

The same girl by the sea with her long ponytail dyed pastel lavender, everything else unchanged

New hair: pastel lavender

Flux Kontext

Change the girl's long black ponytail to pastel lavender hair in the same high ponytail, with slightly darker roots and a soft natural shine. Keep her clothes, the skateboard, the beach behind her and the warm sunlight unchanged, while keeping the same facial features and grin.

A 2x2 collage of the same girl with a buzz cut, curtain bangs, long waves and a slick back, each captioned

Collage: four hairstyles on one face

Nano Banana

A square 1:1 collage made from this photo: a 2x2 grid of four photographs separated by thin white gutters, with a thin white border around the whole grid. Each cell shows the same girl from this photo, head and shoulders, in the same white tee, the same warm beach light and the same relaxed grin, with a different hairstyle. Top left: a real buzz cut, the whole head clipped very short with a uniform clipper guard, about three millimeters all over, the scalp visible through the hair, no length left on top, no fringe and no ponytail. Top right: curtain bangs framing her face with shoulder-length hair. Bottom left: long loose waves past her shoulders. Bottom right: hair slicked straight back with a glossy finish. Her hair stays black in every cell. Under each cell a small caption in plain black sans serif on the white gutter: "BUZZ CUT", "CURTAIN BANGS", "LONG WAVES", "SLICK BACK". Keep her face and features identical in all four cells so she is clearly the same person. Those four captions are the only text, each appearing once, spelled exactly. No logos, no watermark.

2x2 photo collages: the Gemini AI photo trend

These are the Gemini AI photo prompts people share: one upload, four versions of you in a grid, such as four decades, four seasons or four outfits. They also work in the Gemini app; here you can try them on Nano Banana. Tip: give every square a position and an exact caption in quotes, and ask for the same face in all four.

A 2x2 collage of the same man in the 1920s, 1950s, 1970s and 1990s, each square captioned with its decade

Collage: the same man in four decades

Nano Banana

A square 1:1 collage made from this photo: a 2x2 grid of four photographs separated by thin white gutters, with a thin white border around the whole grid. Each cell shows the same man from this photo, head and shoulders, in a different decade, with the period's clothes, hair and photo processing. Top left: the 1920s, a three-piece suit and a flat cap, sepia tone, soft focus. Top right: the 1950s, a tailored grey suit, a thin tie and neatly parted hair, crisp black and white. Bottom left: the 1970s, a wide-collar patterned shirt and a corduroy jacket, a fuller short afro, warm faded color film. Bottom right: the 1990s, an oversized denim jacket over a hoodie, a short fade, bright point-and-shoot flash color. Under each cell a small caption in plain black sans serif on the white gutter: "1920s", "1950s", "1970s", "1990s". Keep his face, beard and features identical in all four cells so he is clearly the same person. Those four captions are the only text, each appearing once, spelled exactly. No logos, no watermark.

A 2x2 collage of the same couple on the train platform in spring, summer, autumn and winter, each captioned

Collage: one platform in four seasons

Nano Banana

A square 1:1 collage made from this photo: a 2x2 grid of four photographs separated by thin white gutters, with a thin white border around the whole grid. Each cell shows the same couple from this photo on the same train platform, framed the same way, in a different season. Top left: spring, light jackets, soft sunshine and blossom petals drifting across the platform. Top right: summer, t-shirts, bright hard sun and a blue sky. Bottom left: autumn, her camel coat and his navy rain jacket, golden leaves on the platform, overcast light. Bottom right: winter, wool hats and scarves, falling snow and frosted train windows. Exactly two people in each cell. Under each cell a small caption in plain black sans serif on the white gutter: "SPRING", "SUMMER", "AUTUMN", "WINTER". Keep both faces identical in all four cells so they are clearly the same two people. Those four captions are the only text, each appearing once, spelled exactly. No logos, no watermark.

A 2x2 collage of the same older woman dressed for the office, the street, an evening out and sport, each captioned

Collage: four outfits for one day

Nano Banana

A square 1:1 collage made from this photo: a 2x2 grid of four photographs separated by thin white gutters, with a thin white border around the whole grid. Each cell shows the same grandmother from this photo, from the knees up, with her grey bun, her glasses on her head and her warm smile, in a different outfit and a simple matching setting. Top left: office, a navy blazer over a cream blouse, by a bright window with a laptop under her arm. Top right: street, a quilted khaki jacket, a striped tee and white sneakers, on a city sidewalk. Bottom left: evening, a deep emerald silk dress with a pearl necklace, in a warm lamplit restaurant. Bottom right: sport, a lilac track jacket, leggings and running shoes, in a green park. Under each cell a small caption in plain black sans serif on the white gutter: "OFFICE", "STREET", "EVENING", "SPORT". Keep her face and features identical in all four cells so she is clearly the same person. Those four captions are the only text, each appearing once, spelled exactly. No logos, no watermark.

Turn a photo into art

Good for a gift print or an avatar from any photo. Tip: name the texture of the medium, like paper grain or fingerprints in clay, and ask to keep the face. Both made in Flux AI.

The grandmother folding dumplings painted as a loose watercolor with soft washes, the kitchen shelves and rice cooker kept

Art style: loose watercolor

Flux Kontext

Change the style of this photo to a loose watercolor painting on textured cold-press paper: soft washes of color bleeding at the edges, fine pencil lines showing through, white paper left for highlights, a gentle warm palette. Keep the older woman with the grey bun folding dumplings, the kitchen and the composition where they are, while keeping the same facial features and warm smile.

The girl with the skateboard remade as a soft clay stop motion figure on the same seaside ledge under palm trees

Art style: clay stop motion

Flux Kontext

Change the style of this photo to a soft clay stop-motion animation look: everything sculpted from matte plasticine with smooth rounded shapes, tiny fingerprint marks and simple handmade textures, warm miniature set lighting and a shallow tabletop depth of field. Keep the composition, the palm trees, the sea, the ledge and the skateboard where they are, and keep the girl with the black ponytail in the same pose and white tee, while keeping the same facial features and grin.

Add text and signs to a photo

Good for a sign, a label or a menu board that was never there. Tip: put every word in quotes, say how many lines, and that nothing else is written. These were made with the ChatGPT image generator.

The same scruffy dog on the sofa, now wearing a small cardboard sign that reads I ATE THE REMOTE

Add a sign: I ATE THE REMOTE

GPT Image

Edit this photo. Add one thing only: a small square of brown cardboard hanging from a loop of string around the dog's neck, resting against his chest, with hand-written black marker letters that read exactly "I ATE THE REMOTE" on two lines ("I ATE THE" above, "REMOTE" below). Uneven but legible capitals, as if a person wrote it quickly. The sign sits naturally on his fur with a soft shadow and matches the afternoon light. Keep the dog, his flipped ear, the sofa, the throw, the tennis ball and everything else exactly the same. Those four words are the only text in the picture, spelled exactly. No watermark.

The same bowl of ramen with a small wooden A-frame chalkboard on the table reading TODAY SHOYU RAMEN MADE BY HAND

Add a sign: a cafe chalkboard

GPT Image

Edit this photo. Add one thing only: a small A-frame wooden table sign at the back left corner of the table, with a black chalkboard face and white chalk lettering on three lines that read exactly "TODAY", "SHOYU RAMEN" and "MADE BY HAND". Neat hand-drawn chalk capitals with a little chalk dust, one small chalk drawing of a steaming bowl under the words. The sign stands naturally on the table with a soft shadow and the same midday window light. Keep the ramen bowl, the toppings, the steam and every other object exactly the same. Those three lines are the only text in the picture, each appearing once and spelled exactly. No watermark.

Product photos from a messy desk

Good for a shop listing from a phone shot. Tip: make two versions, a white background one for the listing and a styled scene for the ad. Both came from one photo in Nano Banana.

A speckled ceramic mug with a blue dipped rim on a cluttered desk among sticky notes, cables, a crumpled receipt and crumbs

Mug on a messy desk, the photo

Nano Banana

An unedited phone photo of a handmade ceramic coffee mug with a speckled oatmeal glaze and a deep blue dipped rim, standing on a cluttered home office desk. Around it: tangled charging cables, yellow sticky notes, a crumpled receipt, a few crumbs, a laptop edge and a pen without its cap. Overhead office light mixed with window light, a slightly yellow cast, a faint shadow of the phone. The mug is in the center, in sharp focus. Exactly one mug. No logos, no readable text, no watermark.

The blue dipped ceramic mug alone on a pure white background with a soft shadow, the desk clutter gone

Product photo: white background listing

Nano Banana

Make a clean listing photo of the mug from this picture. Remove everything except the mug: the cables, sticky notes, receipt, crumbs, laptop and pen. Place the same mug on a seamless pure white background with a soft natural contact shadow under it. Correct the yellow cast so the speckled oatmeal glaze and the deep blue rim show their true colors, and light it with soft even studio light. Keep the exact shape, handle, glaze pattern and rim of the mug, centered and sharp. Exactly one mug. No text, no logos, no watermark.

The same mug steaming on a light oak table beside a croissant, a linen napkin and a sprig of eucalyptus in morning sun

Product photo: lifestyle scene

Nano Banana

Make a lifestyle shop photo of the mug from this picture. Keep the exact same mug, with its speckled oatmeal glaze, deep blue rim and handle, and move it into a calm morning scene: a pale oak table with a folded linen napkin, a small plate with a croissant, a sprig of eucalyptus, soft window light from the left and long gentle shadows, the background softly blurred. Fill the mug with black coffee and a little steam. The mug stays the hero, sharp and in the center. Exactly one mug. No text, no logos, no watermark.

Make a face older or younger

Good for seeing yourself or a parent at another age. Tip: give the target age in years and keep the eye shape and the smile. Both edits are Flux AI.

The same couple on the platform aged about forty years, both grey-haired and still laughing arm in arm

Age a face: forty years older

Flux Kontext

Change the couple to how they would look about forty years older, in their seventies: the woman with curly hair now has soft silver curls and fine laugh lines, the man with short black hair now has thinning grey hair and a grey stubble, both with natural age spots and gentle wrinkles. Keep their coats, the way she holds his arm, their laughter, the train and the platform unchanged, while keeping the same facial features so they are clearly the same two people.

The grandmother in her kitchen shown decades younger, dark hair tied back, the same apron and smile

Age a face: decades younger

Flux Kontext

Change the older woman to how she would have looked at about twenty-five: smooth skin, glossy black hair in the same low bun, no reading glasses, the same warm crinkly smile. Keep the flowered apron, the beige cardigan, the dumplings on the floured board, the kitchen and the morning light unchanged, while keeping the same facial features, eye shape and expression.

How to write a prompt that edits a photo

Start with the one change, then list what must not move: the face, the expression, the pose, the framing and the light. Name the new thing exactly, with its material and color. Keep to one change per prompt; for a second change, run the next prompt on the result. If the person should still look like themselves, say so in plain words at the end.

The guide on how to write AI image prompts has a whole section on editing a photo you uploaded.

Want the photo to move? The guide on how to write AI video prompts turns a still into a short clip.

No photo to start from? The AI image prompts make a picture from words alone.

Questions about editing a photo with a prompt

Short answers, each pointing at the section above that shows it.

Upload a scan or a phone photo of the print and ask for the steps in order: remove the creases, tears and dust, then fix the color, then sharpen the faces. Ask to keep the film grain so it still looks like a photo of its time. The restore section shows a 1970s snapshot and a 1940s portrait.

Yes. Describe the person by what they wear and where they stand, and say what should fill the space, such as the empty platform behind them. Everything else in the photo stays as it was.

Ask for a 2x2 grid made from your photo, name the decade for each square and its position, and put each caption in quotes. Ask for the same face in all four squares. The collage section has the prompt.

Upload the product as it is, mess and all, and ask for the product alone on pure white with a soft shadow, the glaze, shape and color unchanged. A second prompt can put the same product in a styled scene for an ad.

Say what holds the words (a sign, a card, a chalkboard), put every word in quotes, say how many lines there are and that nothing else is written. The text section above was made with the ChatGPT image generator.

Change one thing per prompt and say what stays: the face, the skin tone, the expression and the pose. Run the next change on the result instead of asking for three at once.

A sharp photo with good light on the face, and the face not too small in the frame. One person or one object is easiest; a crowded group gives the model more to get wrong.

Yes. Copy any prompt, change the details to match your photo and keep what you make. Trying one on N33 AI is free to start.

Reviewed by Arthur Down, editor · Updated · How we check facts