PromptClub
Photograph of friends laughing together on a city rooftop at golden hour, caught mid-conversation and completely unposed.

Collection · 22 prompts

22 Candid Portrait Prompts for Gemini, ChatGPT & More

The best portraits are the ones nobody posed for — the laugh mid-sip at a cafe table, the glance over a shoulder on a busy street, the quiet minute by a train window when nobody knew the camera was there. This collection is built for those moments: twenty-two candid portrait prompts ai users can copy into Gemini, ChatGPT, Nano Banana or Midjourney, each one written around an unstaged scene rather than a studio pose. You will find the cafe laugh and the cafe conversation, a street walk caught mid-step and a street crossing in full stride, friends laughing in a park and on a rooftop at golden hour, a market browse among the spices and a first bite at a night-market stall, a beach laugh in sea wind and a couple surprised by a wave, a train-window thinker and a train reader lost in her book, a flour-dusted kitchen laugh and a couple cooking together, a bookshop reach for a high shelf, a rain-soaked laugh, a solo thinker on a park bench, a picnic mid-joke, an autumn park full of falling leaves, a rain-neon street at night and three friends on a hilltop at sunset. Every prompt is written to work from your own photo, so the faces stay yours and only the setting, light and moment change. Pick the prompt that matches your moment, copy it into your model of choice, and change one detail at a time — the setting, the time of day, the outfit — until the portrait feels like a memory, not a photoshoot.

Photograph of a young couple laughing as a wave rushes around their feet on a beach.
ChatGPT
№ 01

The prompt

Turn your photo into a candid beach portrait of a couple caught by a wave, as if a friend snapped it the second the water surprised them both. A young couple stands at the water's edge, laughing as a wave rushes around their ankles, the woman hopping back with her dress lifted, the man reaching for her hand, both caught mid-laugh with wet feet and sandy toes. She wears a flowing coral sundress; he wears a rolled-sleeve linen shirt and shorts, a woven hat in his free hand, both sun-warmed and barefoot. Behind them the beach glows at late afternoon: wet sand mirroring the sky, gentle waves rolling in, a few distant swimmers, a beach umbrella tilted in the breeze, gulls riding the wind overhead. The light is warm and low, golden sun rimming their hair and shoulders, natural skin texture with visible pores and sun freckles, no smoothing, no posed couple stance. Frame a vertical portrait from the knees up, shot at eye level like a 50mm lens, shallow depth of field, the sea melting into golden bokeh, the laughter shared and completely unplanned.

Tip: Ask for the wave's timing — 'rushing around their ankles' — so the couple reacts to something real instead of posing at the water.

Share
Photograph of two friends deep in conversation over coffee at a cafe table.
ChatGPT
№ 02

The prompt

Turn your photo into a candid cafe portrait of a conversation in full flow, as if a friend photographed the table without either of them noticing. Two women sit across a small marble cafe table, one leaning forward mid-sentence with both hands wrapped around her mug, the other laughing with a hand over her mouth, a half-finished latte and a shared slice of cake between them. One wears a rust knit sweater and gold hoops; the other a white blouse under a denim jacket, hair in a low ponytail, a book face-down beside her cup. Behind them the cafe glows warm: hanging plants, a chalkboard menu, a barista steaming milk, sunlight falling in long bars across the wooden floor, a dog asleep under a nearby table. The light is warm and golden, afternoon sun slanting through tall windows, natural skin texture with visible pores and faint freckles, no smoothing, no camera-aware smiles. Frame a vertical portrait from the waist up, shot at eye level like a 50mm lens, shallow depth of field, the cafe melting into warm bokeh, the conversation real and the moment entirely unstaged.

Tip: Ask for the lean-in — one friend mid-sentence, the other laughing — because a conversation needs two different expressions to feel real.

Share
Photograph of a young couple cooking together in a warm kitchen, tasting and laughing.
ChatGPT
№ 03

The prompt

Turn your photo into a candid kitchen portrait of a couple cooking together, as if someone walked in on the best part of the evening. A young couple stands side by side at a kitchen counter, the woman tasting sauce from a wooden spoon with her eyes closed in delight, the man laughing as he reaches past her for the salt, a pan steaming on the stove behind them. She wears a striped apron over a white tee, hair in a loose bun; he wears a grey henley with sleeves rolled up, a tea towel over one shoulder, both with flour on their hands and happy smiles. Behind them the kitchen is warmly alive: chopped vegetables on a board, a pot of herbs on the windowsill, evening light falling across the tiles, a playlist glowing on a phone propped against the fruit bowl. The light is warm and low, golden evening sun through the window, natural skin texture with visible pores, no smoothing, no styled cooking-show pose. Frame a vertical portrait from the waist up, shot at eye level like a 50mm lens, shallow depth of field, the kitchen melting into warm bokeh, the evening easy and entirely unstaged.

Tip: Give each person a job — one tasting, one reaching for the salt — so the couple looks like they are cooking, not modelling.

Share
Photograph of a young man sitting alone on a park bench, lost in thought.
ChatGPT
№ 04

The prompt

Turn your photo into a candid park-bench portrait of a man lost in thought, as if a passer-by glanced over and caught a private minute. A young man sits alone on a wooden park bench, elbows on his knees, hands loosely clasped, eyes on something far across the lawn, a quiet faraway expression, a takeaway coffee cooling beside him on the bench. He wears a navy wool coat over a grey hoodie, dark jeans and boots, a woollen scarf loosely looped, hair ruffled by a light breeze, a paperback tucked into his coat pocket. Behind him the park is calm and green: a wide lawn, bare winter branches overhead, a path curving away, a squirrel on the grass, soft overcast light making every colour gentle. The light is soft and diffused, cool daylight falling evenly across his face, natural skin texture with visible pores and light stubble, no smoothing, no awareness of being watched. Frame a vertical portrait from the waist up, shot at eye level like a 50mm lens, shallow depth of field, the park melting into soft green bokeh, the mood thoughtful and entirely unstaged.

Tip: Keep the gaze off-frame — 'eyes on something far across the lawn' — so the portrait feels observed, not performed.

Share
Photograph of four friends laughing around a picnic blanket in a sunny meadow.
ChatGPT
№ 05

The prompt

Turn your photo into a candid picnic portrait of friends mid-joke, as if someone lifted the camera at the exact second everyone broke. Four friends sprawl on a checkered blanket in a sunny meadow, one laughing so hard she is wiping her eyes, another mid-gesture with a sandwich in hand, a third leaning back on his elbows grinning, the fourth reaching for the lemonade. They wear easy summer clothes: linen shirts, sundresses, denim shorts and sandals, sunglasses pushed up, a straw hat tipped back on one head. Around them the picnic spreads out: a wicker basket, a bowl of strawberries, paper plates, a guitar leaning against a tree, wildflowers nodding in the grass, a kite caught in a distant tree. The light is bright and warm, midday sun softened by a few high clouds, natural skin texture with visible pores and sun freckles, no smoothing, no arranged group pose. Frame a vertical portrait from the waist up, shot at eye level like a 50mm lens, shallow depth of field, the meadow melting into green and gold bokeh, the laughter loud and completely unplanned.

Tip: Give each friend a different action — laughing, gesturing, leaning back — so the group reads as a real picnic, not a posed lineup.

Share
Photograph of a young woman laughing in the rain on a wet street, face tilted up.
ChatGPT
№ 06

The prompt

Turn your photo into a candid rain portrait of pure joy, as if a friend caught her the second the sky opened. A young woman stands in a sudden downpour on a quiet street, face tilted up to the rain, laughing with her eyes closed, arms slightly out from her sides, hair already soaked and clinging to her shoulders, rain beading on her skin. She wears a soaked yellow raincoat open over a white tee, dark jeans and ankle boots splashing in a puddle, a canvas bag slung across her body, raindrops catching the light as they fall. Behind her the street glistens: wet pavement reflecting shop lights, a closed cafe umbrella, a bicycle chained to a railing, rain streaking through warm evening light, a distant figure hurrying under an awning. The light is warm and wet, street lamps glowing through the rain, natural skin texture with visible pores and rain droplets, no smoothing, no posed wet-hair glamour. Frame a vertical portrait from the knees up, shot at eye level like a 50mm lens, shallow depth of field, the street melting into glistening bokeh, the laughter soaked and completely unguarded.

Tip: Let the rain do the work — ask for wet hair and closed eyes, and skip the umbrella so nothing blocks the face or the light.

Share
Photograph of a young man crossing a city street mid-stride, coat flaring.
ChatGPT
№ 07

The prompt

Turn your photo into a candid street-crossing portrait caught mid-stride, as if a friend snapped it from the kerb as he hurried past. A young man crosses a zebra crossing in quick steps, one foot planted and the other swinging forward, coat flaring slightly with the motion, glancing sideways at the traffic light, a leather satchel bouncing against his hip. He wears a camel overcoat over a black turtleneck, dark trousers and polished boots, a watch flashing at his wrist, hair swept back by the wind. Behind him the city moves: a red bus easing past, a cyclist waiting at the light, shopfronts and a newsstand, wet asphalt reflecting the morning sky, a pigeon taking off from a lamppost. The light is crisp morning light, low and bright, catching the edge of his coat and the side of his face, natural skin texture with visible pores and light stubble, no smoothing, no frozen-model pose. Frame a vertical portrait from the knees up, shot at street level like a 35mm lens, slight motion blur in the background, shallow depth of field, the moment hurried and completely unstaged.

Tip: Keep the motion words in — 'one foot planted, the other swinging forward' — so the stride reads as caught, not frozen.

Share
Photograph of a young woman laughing on a beach as sea wind blows her hair across her face.
Gemini
№ 08

The prompt

Turn your photo into a candid beach portrait caught in a gust of sea wind, as if a friend snapped it mid-run along the shore. A young woman stands ankle-deep in the surf, hair whipped across her face by the wind, laughing as she pushes it back with one hand, the other holding her sandals, a wave breaking white around her feet. She wears a flowing white sundress with thin straps, damp at the hem, a shell bracelet on her wrist, no shoes, skin sun-warmed and glowing. Behind her the beach stretches wide: wet sand reflecting the sky, a few distant walkers, a fishing boat on the horizon, gulls wheeling over the water, late-afternoon sun turning the sea to gold. The light is warm and low, rimming her hair and shoulders, natural skin texture with visible pores and sun freckles, no smoothing, no posed beach glamour. Frame a vertical portrait from the knees up, shot at eye level like a 50mm lens, shallow depth of field, the sea melting into golden bokeh, the laughter real and the moment gloriously unplanned.

Tip: Ask for the wind before the dress — 'hair whipped across her face' gives the model motion to work with and keeps the pose from freezing.

Share
Photograph of a young woman reaching for a book on a high shelf in a warm bookshop.
Gemini
№ 09

The prompt

Turn your photo into a candid bookshop portrait caught mid-browse, as if a friend spotted her between the shelves and snapped it quietly. A young woman stands on tiptoe reaching for a book on a high shelf, one hand on the shelf edge for balance, eyes on the spine she wants, a small smile of discovery, two books already tucked under her other arm. She wears a camel-coloured trench coat over a cream sweater, dark jeans and ankle boots, a canvas tote on her shoulder, hair in a loose braid over one shoulder. Behind her the bookshop glows: floor-to-ceiling wooden shelves packed with spines, a brass reading lamp, a ladder leaning against the far wall, a cat asleep on a stack of paperbacks, warm light pooling on the wooden floor. The light is warm and low, falling through tall windows onto the shelves and the side of her face, natural skin texture with visible pores, no smoothing, no camera-aware pose. Frame a vertical portrait from the waist up, shot at eye level like a 50mm lens, shallow depth of field, the shelves melting into warm bokeh, the moment quiet and entirely unposed.

Tip: Ask for the reach — 'standing on tiptoe, hand on the shelf edge' — a body mid-action is the fastest way to kill a posed look.

Share
Photograph of a young woman laughing mid-sip at a cafe table, candid and unposed.
Gemini
№ 10

The prompt

Turn your photo into a candid cafe portrait caught mid-laugh, as if a friend snapped it without warning. She sits at a small wooden table by a tall cafe window, a ceramic cup halfway to her lips, head tilted back in a genuine laugh, eyes crinkled shut, one hand resting loosely on the table beside an open notebook and a half-eaten croissant. She wears a soft cream knit sweater over a white collared shirt, hair loosely pinned up with a few strands falling free, small gold hoops catching the light. Behind her the cafe hums softly: a barista at the espresso machine, warm pendant lights, plants on the windowsill, blurred strangers at the counter, afternoon sun slanting through the glass in long golden bars. The light is warm and natural, falling across her face from the window side, natural skin texture with visible pores and faint freckles, no smoothing, no studio gloss. Frame a vertical portrait from the waist up, shot at eye level like a 50mm lens, shallow depth of field, the background melting into warm bokeh, the moment unposed and alive.

Tip: Ask for the laugh first and the cafe second — if the smile looks staged, add 'mid-laugh, eyes crinkled shut' and regenerate from your original photo.

Share
Photograph of a young woman laughing in a kitchen with flour on her cheek and apron.
Gemini
№ 11

The prompt

Turn your photo into a candid kitchen portrait caught mid-recipe, as if someone walked in just as the flour went everywhere. A young woman stands at a wooden kitchen counter, laughing with her head tilted back, a streak of flour across her cheek and another on her apron, a wooden spoon in one hand and a mixing bowl tucked against her hip. She wears a sage-green apron over a cream tee, sleeves pushed to the elbows, hair in a messy bun with flour-dusted strands escaping, a smear of dough on one knuckle. Behind her the kitchen is warmly lived-in: open shelves of mismatched mugs, a kettle steaming on the stove, a window full of late-morning light, a cat watching from a chair, a tray of half-shaped buns waiting on the counter. The light is soft and warm, falling across her face from the window, natural skin texture with visible pores and faint freckles, no smoothing, no styled food-blog pose. Frame a vertical portrait from the waist up, shot at eye level like a 50mm lens, shallow depth of field, the kitchen melting into warm bokeh, the laughter messy and real.

Tip: Keep the flour on one cheek, not everywhere — one messy detail reads as a real moment, while flour all over reads as a costume.

Share
Photograph of a young woman browsing a colourful spice stall at a market, leaning in curiously.
Gemini
№ 12

The prompt

Turn your photo into a candid market portrait caught while she browses, as if a friend photographed her without her noticing. A young woman leans over a spice stall, one hand lifting a small scoop of deep red chilli powder toward her face, eyes down on the colour, a curious half-smile, a woven basket hooked over her other arm. She wears a rust-coloured linen shirt with the sleeves rolled up, dark trousers and leather sandals, a thin scarf knotted loosely at her neck, hair tied back in a low bun. Behind her the market bursts with colour: mounds of turmeric and dried rose petals, hanging strings of garlic, a vendor weighing lentils, striped awnings overhead, warm shafts of light falling between the stalls. The light is warm and dappled, catching the spice dust in the air and the side of her face, natural skin texture with visible pores, no smoothing, no camera-aware pose. Frame a vertical portrait from the waist up, shot at eye level like a 50mm lens, shallow depth of field, the market melting into colourful bokeh, the moment curious and completely unstaged.

Tip: Name one spice colour plainly — 'deep red chilli powder' — so the market reads as a real stall instead of a generic colourful blur.

Share
Photograph of two friends laughing together on a grassy park slope beneath a tree.
Gemini
№ 13

The prompt

Turn your photo into a candid park portrait of two friends caught mid-laugh, as if someone snapped it during a walk that went on longer than planned. Two young women sit on a grassy slope beneath a wide tree, one leaning back on her hands laughing with her head thrown back, the other mid-gesture telling the story, a picnic blanket and two paper cups scattered beside them. One wears a mustard cardigan over a striped tee and jeans; the other a denim jacket, white top and a floral skirt, sunglasses pushed up into her hair. Behind them the park stretches out in soft green: a winding path, a distant dog walker, dappled light falling through the leaves, a bicycle leaning against a bench. The light is soft late-afternoon gold filtered through the canopy, natural skin texture with visible pores and faint freckles, no smoothing, no arranged poses. Frame a vertical portrait from the waist up, shot at eye level like a 50mm lens, shallow depth of field, the background melting into green bokeh, the laughter genuine and unposed.

Tip: Use a photo where both friends are already laughing — the model holds real laughter better than it invents it from a straight-faced original.

Share
Photograph of three friends laughing on a city rooftop at golden hour.
Gemini
№ 14

The prompt

Turn your photo into a candid rooftop portrait of friends at golden hour, as if someone turned around and caught the group mid-conversation. Three friends sit on the low parapet of a city rooftop, legs dangling, one mid-sentence with both hands in the air, the other two laughing, paper cups and a shared bag of snacks between them on the concrete. They wear easy evening clothes: a denim jacket, a striped sweater, a linen shirt with rolled sleeves, sneakers and sandals, hair moving in the rooftop breeze. Behind them the city glows at golden hour: tower blocks catching the last sun, water tanks and washing lines on neighbouring roofs, a sky grading from amber to soft violet, the first windows lighting up below. The light is warm and low, rimming hair and shoulders in gold, natural skin texture with visible pores, no smoothing, no grouped pose. Frame a vertical portrait from the waist up, shot at eye level like a 50mm lens, shallow depth of field, the skyline melting into golden bokeh, the friendship loud and completely unstaged.

Tip: Name the gesture — 'both hands in the air, mid-sentence' — so the group looks caught talking instead of lined up for a photo.

Share
Photograph of a young man mid-stride on a sunny street, glancing back over his shoulder.
Gemini
№ 15

The prompt

Turn your photo into a candid street portrait caught mid-walk, as if a friend called his name from across the road. A young man strides along a sunlit pavement, caught mid-step with one foot just leaving the ground, head turned over his shoulder toward something off-frame, a half-smile on his face, a canvas tote slung over one shoulder and a paper coffee cup in his hand. He wears a faded olive jacket over a white tee, dark jeans and scuffed sneakers, hair tousled by the breeze. Behind him the street carries on: a cyclist gliding past, a shop awning in faded red, a bus shelter poster, pigeons scattering near a kerb, warm afternoon light raking across the brick facades. The light is low and golden, catching the side of his face and the edge of his jacket, natural skin texture with visible pores and light stubble, no smoothing, no posed stillness. Frame a vertical portrait from the knees up, shot at street level like a 35mm lens, slight motion in the background, shallow depth of field, the moment unstaged and full of life.

Tip: Keep the head-turn in the prompt — a subject looking away from the camera is what sells the candid feel, so never ask for eye contact here.

Share
Photograph of a young man gazing out of a train window at the passing countryside.
Gemini
№ 16

The prompt

Turn your photo into a candid train-window portrait of quiet thinking, as if a fellow passenger glanced across the carriage and caught the moment. A young man sits by a large train window, forehead nearly touching the glass, eyes on the landscape rushing past, a soft thoughtful expression, one hand resting against his chin, a paperback face-down on his knee. He wears a charcoal knit sweater over a checked shirt, dark trousers, a leather-strapped watch, hair slightly flattened on one side from the headrest. Outside the window the countryside blurs by: green fields, a river catching the light, a small station flashing past, soft clouds over low hills, the glass carrying a faint reflection of the carriage lights. The light is soft and overcast, cool daylight washing across his face from the window, natural skin texture with visible pores and light stubble, no smoothing, no awareness of the camera. Frame a vertical portrait from the chest up, shot at eye level like a 50mm lens, shallow depth of field, the world outside melting into motion, the mood still and completely unposed.

Tip: Match your photo's light to the scene — an overcast original blends into a train window far better than a sunny selfie.

Share
Photograph of three friends laughing on a hilltop path in golden sunset light.
Midjourney
№ 17

The prompt

Turn your photo into a candid golden-hour portrait of friends caught in the last light, as if someone turned around at the end of a long walk and snapped the group mid-laugh. Three friends walk along a hilltop path at sunset, one laughing with her head thrown back, another mid-story with a hand on a friend's shoulder, the third grinning at something off-frame, their shadows stretching long across the grass. They wear warm evening clothes: a denim jacket, a knitted sweater, a linen shirt with rolled sleeves, jeans and boots, hair catching the wind, a backpack slung over one shoulder. Behind them the evening opens wide: rolling hills in soft gold, a valley haze below, a sky grading from amber to rose, the sun low and huge near the horizon. The light is warm and dramatic, low sun rimming hair and shoulders in gold, natural skin texture with visible pores, no smoothing, no grouped sunset pose. Frame a vertical portrait from the waist up, shot at eye level like a 50mm lens, shallow depth of field, the hills melting into golden bokeh, the friendship warm and completely unplanned.

Tip: Put the sun behind the group — backlight rims the hair in gold and keeps every face in soft, flattering shadow instead of harsh light.

Share
Photograph of a young man on a rain-slicked neon street at night, looking up from his phone.
Midjourney
№ 18

The prompt

Turn your photo into a candid rain-neon portrait with cinematic street light, as if a friend caught him glancing up from his phone on a wet night. A young man stands on a rain-slicked city street at night, looking up and away from his phone toward the neon signs overhead, rain beading on his jacket, a thoughtful half-smile, one hand in his pocket and the other holding the glowing phone. He wears a black leather jacket over a dark hoodie, black jeans and boots, a beanie, hair damp at the edges, neon pink and blue washing across his face in soft colour. Behind him the night street glows: neon signs reflecting on wet pavement, a late bus passing in a blur of light, a shuttered shopfront, rain streaking through the lamplight, a distant figure with an umbrella. The light is electric and wet, neon colour against deep night shadow, natural skin texture with visible pores and rain droplets, no smoothing, no posed noir stance. Frame a vertical portrait from the waist up, shot at eye level like a 50mm lens, shallow depth of field, the street melting into neon bokeh, the moment cinematic and completely unstaged.

Tip: Let the neon stay in the background and on the edges — if the colour floods the face, ask for 'neon glow behind, face in soft light'.

Share
Photograph of a young woman laughing among falling golden leaves in an autumn park.
Nano Banana
№ 19

The prompt

Turn your photo into a candid autumn portrait caught mid-laugh among falling leaves, as if a friend tossed a handful of leaves just as the camera clicked. A young woman stands on a park path scattered with golden leaves, laughing as a few leaves drift down around her, one hand raised to catch them, the other holding her coat closed, her scarf fluttering in the breeze. She wears a mustard wool coat over a cream sweater, dark jeans and leather boots, a plaid scarf, hair loose and moving in the wind, a leaf caught in her hair. Behind her the park burns with autumn colour: maple trees in gold and rust, a carpet of fallen leaves, a wooden bench, a distant couple walking a dog, low sun slanting through the branches. The light is warm and low, late-afternoon gold filtering through the canopy, natural skin texture with visible pores and faint freckles, no smoothing, no arranged autumn pose. Frame a vertical portrait from the knees up, shot at eye level like a 50mm lens, shallow depth of field, the park melting into golden bokeh, the laughter bright and completely unplanned.

Tip: Ask for leaves in the air, not just on the ground — falling leaves give the frame motion and make the laugh feel timed, not posed.

Share
Photograph of a young man watching the sunset from a city rooftop, hands on the parapet.
Nano Banana
№ 20

The prompt

Turn your photo into a candid rooftop portrait of a man watching the sunset, as if a friend stepped onto the roof and caught him mid-thought. A young man stands at the edge of a city rooftop, hands resting on the low parapet, eyes on the horizon where the sun is sinking, a quiet half-smile, his jacket collar turned up against the evening breeze, a mug of tea cooling beside him on the concrete. He wears a rust-coloured field jacket over a grey tee, dark jeans and boots, a canvas cap in his back pocket, hair moving in the wind. Behind him the city glows at dusk: rooftops and water tanks in silhouette, windows lighting up floor by floor, a sky grading from amber to deep violet, a few early stars overhead. The light is warm and fading, the last sun gilding the side of his face, natural skin texture with visible pores and light stubble, no smoothing, no posed silhouette. Frame a vertical portrait from the waist up, shot at eye level like a 50mm lens, shallow depth of field, the skyline melting into dusk bokeh, the moment still and completely unstaged.

Tip: Ask for the horizon gaze — 'eyes on the sunset' — so the portrait feels like a private minute the camera happened to catch.

Share
Photograph of a young man taking his first delighted bite at a night-market food stall.
Nano Banana
№ 21

The prompt

Turn your photo into a candid street-food portrait caught at the first bite, as if a friend snapped it the second the flavour landed. A young man stands at a night-market food stall, eyes wide and delighted, mid-bite into a steaming skewer, one hand catching a drip, the other holding a paper tray, a grin spreading across his face. He wears a black tee under an open denim shirt, dark jeans and sneakers, a crossbody bag, hair slightly damp from the warm night air, a smear of sauce at the corner of his mouth. Behind him the market hums: glowing stall lights, steam rising from grills, strings of bulbs overhead, a vendor flipping skewers, crowds drifting past in warm blur, signs in bright colours. The light is warm and rich, stall lamps and neon glowing against the night, natural skin texture with visible pores, no smoothing, no posed foodie stance. Frame a vertical portrait from the waist up, shot at eye level like a 50mm lens, shallow depth of field, the market melting into golden bokeh, the delight genuine and completely unplanned.

Tip: Name the reaction, not just the food — 'eyes wide, mid-bite, grin spreading' — because the first-bite face is the whole portrait.

Share
Photograph of a young woman reading a paperback by a train window in evening light.
Nano Banana
№ 22

The prompt

Turn your photo into a candid train portrait of a reader lost in her book, as if a fellow passenger looked up and caught the quietest seat in the carriage. A young woman sits by a train window with a paperback open in both hands, eyes down on the page, a small smile at something she just read, a strand of hair tucked behind her ear, a coffee cup resting on the fold-down tray. She wears a soft green cardigan over a white top, a long skirt, ankle boots, a canvas tote at her feet, a pair of reading glasses pushed up onto her head. Outside the window the evening landscape slides past: fields turning gold, a river catching the last light, scattered houses with lamps coming on, soft clouds over low hills. The light is warm and low, evening sun slanting through the window across the pages and her face, natural skin texture with visible pores and faint freckles, no smoothing, no awareness of the camera. Frame a vertical portrait from the chest up, shot at eye level like a 50mm lens, shallow depth of field, the world outside melting into motion, the mood calm and entirely unposed.

Tip: Keep the book open and the eyes down — a reader looking at the camera breaks the spell, so never ask for eye contact here.

Share

Step by step

How to make candid portraits from these prompts

  1. 1

    Start with a relaxed photo, not a posed one. Candid prompts work best from an original where you are already at ease — a smiling selfie, a photo a friend took while you were talking, a shot where you look away. A stiff, straight-faced photo fights the unposed energy these prompts are written for, so pick your most natural picture first. 2. Copy the prompt and change one detail first. The easiest first change is the setting or the outfit — move the cafe laugh to a park bench, swap the raincoat colour, change the bookshop for a market. One change per run keeps the result predictable; changing the setting, light and action at once hides what went wrong. 3. If the result looks posed, add motion and look away. Ask for 'looking away from the camera' and name one action — mid-laugh, mid-stride, mid-sentence, reaching for something. A body doing something cannot pose, and a gaze aimed anywhere but the lens is the single strongest candid signal you can give a model. Regenerate from the original, not the stiff version. 4. If the face drifts, shorten the prompt. Long prompts give the model more room to reinvent the face. Re-upload the original photo, paste a trimmed version of the prompt, and add a plain instruction to keep the same face and skin tone. Gemini and ChatGPT hold faces best in this collection; Midjourney suits the cinematic rain-neon and golden-hour scenes, where light is the star. 5. If the hands look wrong, simplify the pose. Ask for hands resting flat, holding one simple object, or tucked into pockets — the coffee cup, the book and the mixing bowl in these prompts are there for exactly that reason. Avoid interlocked fingers and small objects, and re-run with the same simple pose rather than adding hand detail.

Questions, answered

Frequently Asked Questions

Which model should I start with for candid portraits?

Start with Gemini — nine of this collection's twenty-two prompts were written for it, including the cafe laugh, the market browse and the bookshop reach, and it is the strongest at keeping your face recognisable while rebuilding the scene around you. ChatGPT is the easiest model to adjust if a result feels too posed: just ask for 'more candid, looking away' and regenerate. Nano Banana suits the night-market and rooftop prompts, and Midjourney gives the most cinematic light for the rain-neon and golden-hour friends portraits.

Can I use my own photo with these candid prompts?

Yes — that is how they are meant to be used. Upload a clear photo of yourself, then paste the prompt alongside it. The model keeps your face and rebuilds the setting, the light and the moment around you. A photo where you are already relaxed works best: a smiling selfie drops naturally into the cafe laugh, while a stiff, straight-faced photo fights the candid energy the prompts are written for. If your photo is dark or blurry, the model has too little to hold on to and the face drifts.

Why do my candid portraits still look posed?

Ask for motion and a gaze that is anywhere but the camera. The prompts are already written that way — laughing mid-sip, glancing over a shoulder, reaching for a book, tasting sauce — because a body doing something cannot pose. If a result still looks staged, add 'looking away from the camera' and name one action, such as 'mid-laugh' or 'mid-stride', then regenerate from your original photo rather than editing the stiff version.

Can I change the setting — move a cafe portrait to a park?

Yes, and it is the easiest change in this collection. Swap the setting and keep the action: move the cafe laugh to a park bench, the market browse to a bookshop, the beach laugh to a rooftop at golden hour. Keep the light description and the gesture, since those are what make the portrait feel candid. Change one element per run so you can see what each edit does, and name the new setting plainly.

What should I do if the hands look wrong?

Simplify the hands. Ask for hands resting flat, holding one simple object, or tucked into pockets — the prompts already use those poses, which is why the coffee cup, the book and the mixing bowl appear so often. Interlocked fingers and hands holding small objects are the hardest poses for image models. If a run still tangles the fingers, re-run with the same simple pose rather than adding more hand detail.

Can I use the portraits I make commercially?

For personal use — your own profile pictures, prints for the home, sharing a moment with friends — that is what these prompts are for. For commercial use, such as a photographer's portfolio, a brand's social feed or anything you sell, check the terms of the AI platform that generated the image, since commercial rights vary by platform and plan. We are not giving legal advice; the platform's own terms are the authority.

Keep browsing

More collections