by ByteDance

Seedance 2.5 Prompts

Seedance 2.5 is a next-gen AI video model that generates continuous, synchronized 30-second 4K clips from a single prompt.

video

FORMAT 20 seconds, vertical 9:16, real-time speed, one continuous uncut take. No slow motion, no speed ramps, no cuts. Photorealistic live-action action-comedy. No on-screen text, no camcorder date stamp, no REC overlay, no captions, no music. STYLE: authentic amateur footage shot on a consumer MiniDV tape camcorder, handheld by a café customer. Everything in front of the lens is completely real and richly detailed: real people, real skin, real cobblestones, real sheet metal, real fruit, real physics. A genuine home video of a real, unstaged event. REFERENCES Image #1 defines only the robot: a clearly adult woman in her mid-twenties. Preserve exactly: soft adult face, brown eyes, natural brows, subtle makeup; long straight light brown hair with a center part falling past the shoulders; small stud earrings; slim build with long legs; fitted beige short-sleeved T-shirt with a small dark chest logo with no legible text, hem ending just above the waist; high-waisted dark purple cycling shorts with a small front drawstring; white chunky running sneakers. Exposed mechanical waist between shirt hem and waistband: brushed silver mechanical blocks and actuators, small bright white LED lights, bundles of blue, yellow and black wires, metal brackets, one blue cable running down her left hip. Her torso always connects to her pelvis through this mechanism, same scale from every angle. Face, arms and legs have real human skin: pores, fine hairs, natural sheen. Do not copy the photo's background, people or floor. REALISM The viewer must believe this was filmed in real life. Ordinary people with asymmetrical faces, uneven skin, sweat, messy hair, worn creased clothing. Real hard sunlight, bounce light from pale walls, dust in sunbeams. Real mass: the van is slow to start, accelerates steadily under gravity the whole way down, and is frighteningly heavy and fast at the bottom; light objects react instantly. Every contact has a visible cause and consequence. People flinch, hesitate, collide while fleeing. Nothing happens in perfect sync. Physics rule: the robot is far heavier and stronger than she looks. This is the only supernatural element and it is consistent: she cannot stop the van instantly, it shoves her backward ten meters, her sneakers grind and smoke, her waist mechanism compresses hard, the hood deforms around her palms. Everything else obeys real physics: a three-ton van rolling downhill under gravity, suspension bouncing on cobblestones, crate inertia, oranges rolling individually. Scale: the van is about 2.3 meters tall and clearly dwarfs her. CHARACTERS The Robot: calm, silent, unreadable. Effort shows only in her stance, body angle and mechanism, never in a grimace. No bulging muscles. The Driver: mid-forties, heavy-set, sunburned, grey stubble, faded green work polo, cargo shorts, work boots, flat cap. Café customers: about six adults and a waiter in a black apron, ordinary clothes, real faces. The Van: parked facing downhill from the first frame to the last, front always toward the camera. An old white box delivery van, scratched, rear double doors, no lettering, loaded with open wooden crates of oranges. ENVIRONMENT Sunny morning, old European town. A visibly steep straight cobblestone street, about a 15 percent slope, roughly 40 meters long: the facades step down the hill, doorways have steps on one side, parked scooters lean on their stands, the slope is obvious in every shot. At the bottom the street opens into a completely flat small piazza junction about 15 meters across. On the far side of the junction, slightly to the right of the street's line, a café terrace: round metal tables, wicker chairs, white cups. Two empty wicker chairs and a small table stand out in the open junction in front of the terrace. Pastel facades, wooden shutters, flower boxes. Low warm sun from the upper left. No legible signs, plates or logos. TIMELINE 00:00-00:01.5 | CALM, ONE BREATH Wide handheld view up the steep street. The robot sits in profile at the next table, sipping an espresso, waist LEDs glowing. At the top of the hill the white van is parked facing downhill: its front faces the camera, its rear double doors at the uphill end stand open. The Driver has just pulled a crate of oranges out of the back and is carrying it toward a doorway beside the van. 00:01.5-00:04.5 | IT BREAKS AWAY A metallic clunk: the handbrake was not set. The van breaks away and starts rolling forward, nose first, straight down the hill toward the camera. It never turns around. The Driver drops his crate, oranges burst across the stones, he lunges, grabs the open rear door, is dragged three steps and falls flat on the stones, shouting "No, no, NO!". The van gains speed, suspension bouncing on the stones. Rear doors bang open and shut, crates topple, a stream of oranges pours out behind it. The operator jerks a clumsy zoom toward the van, then zooms back out. 00:04.5-00:07.5 | IT'S COMING The van keeps accelerating, now about 35 km per hour, heavy, wobbling on worn suspension. Its right side catches a parked scooter and drags it along the stones for several meters with a screech and a shower of real metal sparks, then the scooter spins away into a doorway. Pedestrians flatten against walls. On the terrace customers scream and scatter, chairs topple, a cup shatters. The waiter shouts "Get out of the way!". At 00:05.5 the robot sets her cup down, stands, takes four quick steps into the center of the flat junction, and stops. She does not walk toward the van. She turns to face it, plants her feet wide, one foot slightly back, bends her knees, raises both palms at hood height. She waits. 00:07.5-00:09 | LAST SECOND The van leaves the slope and bursts onto the flat junction, filling the left half of the frame, engine silent, tires roaring on stone. A man screams "Lady, MOVE!". 00:09-00:11.5 | I

you find businesses that sell outdoor spaces from drawings and turn their 2d landscape plans into photoreal walkthrough videos with a live blueprint tracker in the corner. you handle the lead research, the sample and the build. every frame is judged against the client's own plan, and nothing ships until the garden in the video matches the drawing. - niche: [landscapers / garden designers / pool builders / decking and patio companies / outdoor kitchen installers] - area: [county or city] - offer: a free 10 second walkthrough of one of their own plans, then $[PRICE] per plan - my details: [name, email, booking link] - i'm away and won't answer questions. make the calls yourself and keep going, but never send anything without my approval. - use a google maps scraper to pull every [niche] in [area] with a website - crawl each site and keep the ones with 2d plans, design drawings or before/after projects on their pages but no 3d walkthroughs or video - for each one, save their best plan, the owner or sales contact, and a one line note on how they sell today - only use public pages and plain requests. never work around bot checks or logins - redraw their plan as a clean top-down cad blueprint: thin white linework on deep navy, labels for every area, north arrow, scale bar and dimensions - pick where the camera starts (usually the back door or terrace) and work out what sits on the camera's left and right from there. facing away from the house, the plan's left side is on the camera's right. get this right before any render - render one photoreal start frame with gpt image 2.6, using the blueprint as reference, from that exact spot at eye height - real photo look: soft natural daylight, neutral white balance, no sunset orange, no lens flare, real materials with weathering, varied planting, nothing cloned or perfectly symmetrical - draft a short personal email and dm with the sample and the offer. no invented prices, numbers or claims - put business, plan, sample, contact and drafts in a sheet and show me before anything is sent once a client pays: - check the credit cost of every generation first and tell me the total before spending - test the camera move on a cheap 5 second model first. only run the final once the motion holds - final clip: 10 seconds, 16:9, seedance 2.5 or kling 3.0 from the start frame - camera: one continuous gimbal move at a brisk walking pace, no cuts. a new area comes into view every 2 to 3 seconds. slow push-ins look dead, so keep it moving - the layout, materials and lighting must not morph during turns - pull frames every half second and map the camera's real path onto the blueprint: position, direction faced, and timing - build the tracker in code with python and ffmpeg: blueprint in a rounded panel top left, a glowing gold dot on the camera's position, a soft light cone showing where it faces, a dashed trail behind it, and each area lighting up when the camera looks at it - time the dot to the actual frames, not to the prompt. if the camera turns at 3 seconds, the dot turns at 3 seconds - the deliverable is a 1080p mp4 that plays on a phone the builder never grades its own work. - after the start frame and after the final video, send frames plus the blueprint to a fresh critic that has seen none of the work - it answers yes or no, at most five problems ranked, each with a concrete fix, and what is already good - it checks: does the garden match the plan left to right, does anything warp or melt, does it look like a real photo or like ai, is the pace fast enough, does the dot match the camera - fix the worst first, render again, send to a new critic. repeat until yes - for choices of taste, show a critic both versions labelled a and b and let it pick - the lead sheet with samples and drafted messages, for my approval - per paid client: the final video with the tracker, the blueprint, the start frame, the critic log, credits spent, and a note on anything a human should still check

STYLE: authentic vertical smartphone video filmed handheld by an ordinary passerby on a city sidewalk in the evening. Real phone footage, not cinematic: auto exposure adjusting between bright shop windows and the dark sky, slight digital noise in shadows, mild motion blur on fast movement, autofocus occasionally hunting, slightly crushed blacks, warm sodium streetlight mixed with cool shop-window LED light. Underneath this texture everything is fully real: real people, real skin, real asphalt, real bricks, real dust, real physics. FORMAT 15 seconds, vertical 9:16, real-time speed, one continuous uncut take, no slow motion, no speed ramps. Photorealistic live-action rescue drama with a dry comic ending. No text, captions, UI overlays, timestamp or music. REFERENCE Image #1 defines only the adult female humanoid robot. Preserve exactly: soft youthful adult face, brown eyes, natural brows, subtle makeup; long straight light brown hair with a center part falling past the shoulders; small stud earrings; slim adult build with long legs; fitted beige short-sleeved T-shirt with a small dark chest logo, hem ending just above the waist; high-waisted dark purple cycling shorts with a small front drawstring; white chunky running sneakers. Exposed mechanical waist between shirt hem and waistband: brushed silver mechanical blocks and actuators, small bright white LED lights, bundles of blue, yellow and black wires, metal brackets and fasteners, one blue cable running down her left hip. Her torso is always connected to her pelvis through this mechanism, same scale and construction from every angle. Do not copy the photo's background, crowd, lanyards, stanchions or floor. PHYSICS RULE The robot is far stronger and heavier than she looks. This is the only supernatural element and it is consistent: she absorbs the falling load by bending her knees deeply, her sneakers press into the asphalt and crack it, her waist mechanism visibly compresses and trembles under the weight. Everything else obeys real physics: the load swings and falls with real gravity, bricks are heavy and shatter on impact, dust spreads and settles, the man falls with real weight, liquid spills naturally. CAMERA OPERATOR The operator is never visible, not even in reflections in shop windows. He stands at the entrance of a small coffee shop about four meters from the action, filming at chest height. The camera looks along the sidewalk: the construction scaffolding runs along the right side of the frame, the street with passing cars is on the left, the robot is in the center. Real handheld phone behavior: micro-shake, breathing sway, a slight tilt, subjects drifting off-center, late autofocus. He reacts physically: tilts up fast toward the crack above, tilts back down, flinches and steps back at the impact, then slowly steps forward. These imperfections never hide the key moments: the strap snapping, the pallet tilting, the robot's dash, the push, the pallet falling, the catch, the lowering, the cup being crushed, the lift to his feet and all lines. CHARACTERS The Man: around thirty, slim build, short dark brown hair, clean-shaven, navy bomber jacket over a white T-shirt, black trousers, white sneakers, white over-ear headphones on his head, phone in his left hand, paper takeaway coffee cup with a plastic lid in his right hand. Absorbed in his phone, oblivious. The Worker: construction worker on the scaffolding at third-floor level, about nine meters up, orange hi-vis vest, white hard hat, work gloves. Passersby: four or five adults: a woman with a shopping bag, an older man, a young couple. Ordinary clothing, real faces, all several meters away from the load. Robot: calm, focused, silent. Her face stays composed during the whole rescue; effort shows only in her body and mechanism. No bulging muscles, no grimace. The Man's arc: oblivious, shock, breathless gratitude, deadpan loss. Passersby's arc: idle, frozen fear, stunned silence. LOCATION AND BLOCKING Evening city street just after rain. Wet dark asphalt sidewalk reflecting streetlights and shop windows, shallow puddles. Along the right side: a building under construction with steel scaffolding, green safety netting and plywood hoarding at street level. A tower crane boom crosses the dark sky above with small red warning lights. A pallet of red bricks hangs from the crane hook about twelve meters above the sidewalk on two lifting straps. The bricks are shrink-wrapped in clear plastic and held with steel bands on a wooden pallet, roughly one ton. One top corner of the wrap is already torn, with a few loose bricks on top. Fixed positions: the robot stands by a lit shop window in the center of the frame, about two meters from the Man's path. The Man walks along the sidewalk from the background toward the camera and stops directly under the hanging load to read something on his phone. Passersby are scattered behind and around them. TIMELINE 00:00-00:02.2 | CALM Quiet evening sidewalk. The robot stands by the shop window, waist LEDs glowing softly, reflected in the wet asphalt. The operator films her curiously, focus pulling onto her mechanism. The Man walks into frame, headphones on, scrolling his phone, sipping coffee, and stops under the load. Faint crane motor hum above, traffic noise. 00:02.2-00:04.2 | THE SNAP At 00:02.2 a sharp metallic crack sounds above: one lifting strap snaps. The pallet jerks and tilts steeply, hanging from the remaining strap, swinging. Three loose bricks slide off the torn corner and fall, smashing on the sidewalk two meters to the side of the Man; he does not notice, his headphones block the sound. The operator tilts up fast, catching the tilted swinging pallet against the dark sky, then tilts back down to the street. At 00:02.8 the Worker leans over the scaffold rail and shouts: "Look out! Move!" At 00:03 the robot's head turns up toward the load. She

Cinematic slow-motion wuxia fantasy scene in an ancient maple forest with glowing amber autumn leaves falling through dramatic sunbeams. A graceful Chinese swordswoman with elegant dark hair and ornate hairpins, dressed in flowing white traditional hanfu robes, wields a long sword with fluid, powerful martial arts choreography. Swirling gusts of wind lift autumn leaves into the air, transforming them into a magnificent, coiled spectral dragon made of swirling orange maple leaves that breathes a burst of fire. Atmospheric mist, volumetric backlighting, ethereal glow, cinematic depth of field, photorealistic textures, 8k resolution, smooth camera tracking, no digital watermarks. --ar 9:16

cinematic trailer shot, a fierce blonde-haired woman with glowing purple eyes and dark claws standing atop shipping containers at a foggy cargo harbor at night, facing a massive approaching horde of zombies. She leaps down, unleashing swirling waves of luminous purple magic, fighting her way through enemies with agile combat, acrobatic flips, and powerful shockwaves. Epic dark fantasy action style, moody cinematic lighting, wet reflections, high-end VFX, 4k.

Create a 30-second ultra-realistic cinematic fantasy video. A young woman wearing a black cap, black oversized hoodie, black pants, and white sneakers runs through a mysterious foggy forest. Birds suddenly fly away as she discovers a gigantic ancient dinosaur lying camouflaged beneath the grassy hills. She approaches carefully, climbs onto its massive textured body, and stands on its head as the enormous dinosaur slowly raises its long neck. Epic cinematic wide shots, realistic skin and textures, atmospheric mist, dramatic camera movements, natural overcast lighting, breathtaking scale, Hollywood-level visual effects, photorealistic 4K quality.

Create a beautiful 30-second cinematic 3D animated romantic short film set in a peaceful autumn park. A young woman with long dark-brown hair wearing a stylish lavender dress sits upset on a wooden bench. A handsome young man in a beige long coat approaches her, hoping to make things right. He surprises her with a gorgeous bouquet of red and white roses. At first, she ignores his gesture, but slowly her expression softens. She finally accepts the flowers with a sweet smile, and they share a warm romantic moment. In the final scene, she stands up and hugs him affectionately while holding the bouquet. Golden autumn leaves gently fall around them as the camera pulls back to reveal the beautiful park. Use premium Disney/Pixar-inspired 3D animation, expressive facial emotions, consistent characters, realistic movements, cinematic close-ups, warm golden-hour lighting, smooth transitions, and beautiful romantic storytelling. 16:9 aspect ratio, 4K quality. No subtitles, text, or watermark.

SUBWAY KATANA MASSACRE 15 SECONDS | 16:9 | 60FPS | 8K LOOK Ultra-photorealistic Hollywood live-action martial arts fight inside an underground subway station. CHARACTER LOCK: @image1 is the only protagonist. Preserve exact reference face, hairstyle, beard, body proportions and black tactical outfit. Always completely EMPTY-HANDED. No weapons in his hands. All enemies carry Japanese katanas. Maintain identity throughout. ENVIRONMENT: One underground subway station with long platforms, railway tracks, tiled walls, concrete pillars, metal benches, ticket machines, fluorescent lights, signs and one stationary train. Cool lighting, reflective floors and subtle dust. No other locations. ACTION — NONSTOP COMBAT: 0–3s: Several katana attackers surround @image1. One attacks with a horizontal slash. @image1 ducks underneath, delivers rapid Wing Chun punches to his face and torso, then a spinning back kick to another enemy's chest. 3–6s: Three enemies attack simultaneously. @image1 evades their blades, blocks one sword arm, delivers an elbow to the face and a Muay Thai knee to the torso, then performs a fast suplex takedown on another attacker. 6–9s: Two enemies attack from opposite directions. @image1 ducks one slash, pivots and delivers a jumping knee to the second enemy's face. Follows with rapid punches and a powerful grappling throw. 9–12s: Remaining enemies charge. @image1 evades a katana slash, lands a spinning elbow, sweeps another attacker, then performs rapid punches, a jumping side kick to the chest and an explosive takedown. 12–15s: The final leader attacks with his katana. @image1 evades the blade, delivers rapid punches, a powerful Muay Thai knee and a forward side kick to his chest. The leader falls. @image1 advances toward remaining enemies. Camera rapidly pushes in. Hard cut during combat. STRICT TARGET LOCK: Every attack visibly connects with a human opponent. Punches target faces or torsos. Kicks, knees and elbows hit enemies. Grapples require direct contact. Every katana attack targets @image1, who visibly evades every blade. Never attack empty space, furniture, walls or pillars. Never grab weapons. CAMERA: Dynamic handheld-Steadicam tracking, wide shots, low angles, close-up impacts, lateral tracking, whip pans and fast push-ins. Keep @image1 and his target visible. Realistic impact shake and continuous fight coverage. VISUAL STYLE: Photorealistic Hollywood action film, realistic skin, natural expressions, authentic clothing physics, cool fluorescent lighting, HDR, subtle Kodak 35mm grain. No artificial CGI appearance. MOTION: 2.3x action pacing. Wing Chun punches, Muay Thai knees, side kicks, jumping knees, spinning elbows, evasive footwork and grappling takedowns. Immediate transitions. No pauses, idle poses, slow motion or repeated animations. NEGATIVE: Weapons in protagonist's hands, katana in @image1's hands, guns, knives, missed attacks, empty-space strikes, attacking objects, slow motion, pauses, stiff choreography, excessive spinning, floating bodies, unrealistic physics, duplicate protagonist, identity changes, outfit changes, distorted face, bad anatomy, broken limbs, blood, gore, restaurant, warehouse, shopping mall, outdoor streets, anime, cartoon, video-game graphics, CGI. CONTINUITY: One protagonist only. Exact @image1 identity, hairstyle, beard, outfit and proportions. Enemies always carry katanas. Same subway architecture, lighting and train position throughout. PRIORITY: NONSTOP FAST COMBAT + STRICT HUMAN TARGET LOCK + EMPTY-HANDED @IMAGE1 VS KATANA ENEMIES.

Cinematic 4K shot a powerful dark fantasy woman wearing intricate black warrior attire standing on a stormy beach with dark dramatic clouds and crashing ocean waves a herd of horses gallops past her on the sandy shore Dark magical energy swirls around her hands with crackling blue lightning Massive detailed black feathered wings erupt from her back, and she lifts off the ground, soaring gracefully away over the coastline. Epic atmosphere, photorealistic, smooth cinematic motion.

Photorealistic Hollywood supernatural horror short film, shot on 35mm anamorphic lenses with natural film grain, shallow depth of field and practical lighting. Opens with an instant scare, no establishing shot. Setting: an isolated lakeside cabin at night in heavy rain and thunderstorm, rustic wood interior, one flickering table lamp, dying fireplace embers, a cellar door in the kitchen. CHARACTER LOCK: Claire is a woman in her late 20s with shoulder-length auburn hair, an oversized gray knit sweater, jeans and dark socks, identical in every shot. A framed photo of her late mother sits on the mantel. The creature is original: a very tall, gaunt, gray-skinned humanoid with unnaturally long arms and fingers, joints that bend slightly wrong, long wet black hair hanging over a smooth blank face with only a thin dark seam where a mouth should be. It moves in quick, stuttering bursts and mimics human voices perfectly. Practical-effects look. No blood, no gore. Cold blue storm light against warm amber lamp glow, natural diegetic sound, low string drone from the first frame. No text overlay, no watermark. [0-1s] Hard cut: the cellar door in the kitchen bangs open on its own, rain hammering the windows, the lamp flickering. [1-2s] Close-up: Claire freezes on the sofa, her mug slipping from her hand and shattering on the floorboards. [2-3s] From the dark cellar, her late mother’s warm voice calls softly: “Claire? Sweetheart, come down here.” [3-4s] Extreme close-up: Claire’s eyes fill with tears and dread, the mother’s photograph blurred behind her. [4-5s] Slow dolly shot down the hallway toward the cellar as she grabs a flashlight from a hook. [5-6s] Over-the-shoulder: she steps onto the first stair, wood creaking, the beam trembling. [6-7s] POV down the steps: the flashlight finds wet footprints on the concrete, far too long and narrow to be human. [7-8s] Close-up: she whispers, “Mom?” and her own voice answers “Mom?” from right behind her. [8-9s] Whip-pan: nothing behind her but the empty stairwell. [9-10s] Wide shot: in the far corner of the cellar, a tall gaunt figure stands with its head bowed, hair hanging. [10-11s] The flashlight flickers and the figure is suddenly halfway across the room. [11-12s] Close-up: its head lifts, the hair parting to show the smooth blank face and the thin dark seam. The mother’s voice comes from it: “Claire.” [12-13s] Handheld: she bolts up the stairs, long fingers scraping the steps behind her. [13-14s] Low-angle: she bursts into the kitchen, spins and slams the cellar door, throwing her whole weight against it. [14-15s] Close-up: three heavy thuds bow the wood, and the mother’s voice sobs through the gap: “Claire, please, I’m cold.” [15-16s] Close-up: tears on her face as her trembling hand slides the deadbolt home. [16-17s] She drags the heavy oak kitchen table across the floor against the door. [17-18s] The thuds stop. Silence except for rain, her ragged breathing, and a distant thunder roll. [18-19s] Wide shot: she backs away from the door, listening. [19-20s] Lightning flash: through the kitchen window, the exterior cellar hatch in the yard stands flung open and empty. [20-21s] Close-up: her eyes widen as she realizes it never needed the door. [21-22s] She grabs her car keys and sprints for the front door. [22-23s] Tracking shot: she runs across the muddy yard in the rain toward her car, headlights blinding the lens. [23-24s] Interior car: she locks the doors and fumbles the key into the ignition, the engine roaring to life. [24-25s] The car lurches down the dirt road, the cabin shrinking in the rearview mirror. [25-26s] Close-up: wipers sweeping, radio static, her breath slowing into a tearful, relieved laugh. [26-27s] Close-up in the rearview mirror: the back seat is empty. She exhales. [27-28s] The radio static swells and a soft voice whispers from the back seat in her mother’s tone: “Don’t leave me again.” [28-29s] Extreme close-up: her eyes lock on the mirror as a long, pale hand slowly settles on the headrest behind her. [29-30s] Smash cut to black with a hard sound sting. Audio: storm rain and thunder, creaking wood, a low string drone and heartbeat pulses that drop to near-silence in the quiet beats, mimicked voices with a faint unnatural echo, car engine and wipers, a sharp stinger on the final cut. Camera: slow dolly pushes, whip-pans, low angles, handheld on the chase, extreme close-ups. Realistic shake only on impacts.

20-second cinematic 3D animated comedy short film with high-quality Pixar-style visuals. Scene 1 (0–4 seconds): A beautiful pregnant young woman with curly brown hair, wearing a lavender cardigan, white top, and blue jeans, stands inside a modern home. She gently holds her baby bump with a loving expression. Scene 2 (4–9 seconds): A cute little toddler girl wearing a pink outfit plays basketball inside a colorful indoor basketball court. She picks up a basketball, runs toward a miniature basketball hoop, and throws the ball toward the basket. Scene 3 (9–13 seconds): Show the basketball action from a dramatic low-angle camera perspective. The little girl misses the shot, and the ball bounces away. Cut to the pregnant woman looking shocked and surprised. Scene 4 (13–16 seconds): A handsome young father with curly dark hair, a short beard, and a green jacket appears beside the pregnant woman. Both look toward their little daughter with surprised expressions. Scene 5 (16–20 seconds): The little girl is now lying peacefully on the basketball court, hugging the basketball like a pillow and falling asleep. End with a funny, adorable cinematic moment. Visual Style: High-end Pixar-inspired 3D animation, adorable character designs, expressive facial emotions, realistic fabric textures, cinematic lighting, smooth character animation, vibrant colors, detailed environments, professional camera movements, consistent character appearance throughout the video. Aspect Ratio: 9:16 vertical. Duration: 20 seconds. Quality: Ultra-detailed 3D animated cinematic short film.

Recreate Video1, replace only the original character with Image1. Strictly preserve 100% of Image1's face, hair, body, clothing and details. Match Video1's movements, expressions, camera angles, framing, lighting, background, timing and pacing. Seamless, photorealistic character replacement. No other changes.

Create an ultra-realistic early-2000s Korean MiniDV vlog featuring a Korean woman in her early 20s visiting a tiny neighborhood nail salon in 2004. She has shoulder-length dark brown hair, natural makeup, a pastel cardigan, and a knee-length skirt. Shot 1 (0–3s): Handheld footage follows her entering the small salon, sitting at a manicure table, browsing vintage nail polish bottles, and choosing a soft-pink shade. Shot 2 (3–5s): Close-up of her selecting the polish bottle while the technician prepares the manicure. She casually flips through an old fashion magazine. Shot 3 (5–8s): Extreme close-up of the technician carefully filing and shaping her natural, unpainted fingernails. Shot 4 (8–12s): Macro shots of glossy soft-pink polish being applied evenly to each nail. She watches with curiosity and slightly wiggles her fingers. Shot 5 (12–15s): She raises both hands toward the camcorder, shows off her finished glossy nails, turns her wrists to catch the light, and smiles cheerfully. Visual & Audio Style: Authentic 2004 South Korean neighborhood salon, consumer MiniDV footage with 4:3 source-video characteristics inside a 16:9 frame. Handheld shake, imperfect autofocus, soft tape detail, faded colors, warm fluorescent lighting, and natural exposure fluctuations. Include realistic filing sounds, polish bottle clicks, magazine pages, quiet conversation, and faint street ambience. Avoid polished commercial-style cinematography. Continuity: Maintain identical character appearance, clothing, salon, technician, and nail polish shade throughout. Show natural progression from unpainted to filed to freshly painted nails. Only two anatomically correct hands with five fingers each. Overall Feeling: An ordinary, charming afternoon in South Korea in 2004, captured as a nostalgic personal home-video memory.

Cinematic suspense-action short film, opens with an instant hook, grounded realistic style, maximum dynamic camera work. Setting: a busy supermarket in the afternoon, long aisles, stocked shelves, shoppers with carts, a checkout lane, flickering fluorescent lights. CHARACTER LOCK: the main character is a man in his early 30s, short dark hair, gray hoodie, dark jeans, identical in every shot. Muted natural color grading, natural sound, tense score from the first frame. No text overlay, no watermark. > > **[0-1s]** Hard cut: a shopper in the cereal aisle suddenly lunges at the person beside him, both tumbling into a shelf. > **[1-2s]** Low-angle shot: boxes and cans scatter across the tiles as nearby shoppers gasp and back away. > **[2-3s]** Close-up: the two figures rise slowly, skin pale, eyes glazed and unfocused. > **[3-4s]** Overhead security-camera angle: shoppers stampede toward the front doors, carts crashing into displays. > **[4-5s]** Tracking shot: several shoppers stumble and fall behind the deli counter, out of frame. > **[5-6s]** Whip-pan to the checkout lane: the pale figures spill over the counter as the cashier ducks out of view. > **[6-7s]** Wide shot: the aisles are littered with dropped baskets and bags, and the group of pale figures has grown to six. > **[7-8s]** Close-up: the main character crouches behind an endcap, breathing hard, watching. > **[8-9s]** He grabs a metal stock pole from a pallet jack, gripping it with both hands. > **[9-10s]** A young cashier hides under a checkout counter as two pale figures approach. > **[10-11s]** He sprints across the floor and shoves a loaded shopping cart into the two figures, pinning them against a shelf. > **[11-12s]** Handheld close-up: he knocks the first one down with a firm swing and it stays on the floor. > **[12-13s]** The second pushes forward; he sidesteps and sends it sprawling into a stack of water bottles. > **[13-14s]** Low-angle shot: he helps the cashier up and points toward the stockroom door. > **[14-15s]** Tracking shot: she runs for the stockroom and slips inside as the remaining figures turn toward him. > **[15-16s]** Wide shot: four pale figures charge down the main aisle, shelves rattling. > **[16-17s]** He backs into the freezer aisle, glass doors fogging, pole raised. > **[17-18s]** He swings a freezer door open into the lead figure, knocking it flat on the icy floor. > **[18-19s]** Close-up: breath fogging, jaw set, he plants his feet as the other three close in. > **[19-21s]** SLOW MOTION: he spins and knocks the nearest one down, then catches the second mid-lunge and slams it onto the tiles. > **[21-23s]** Overhead shot: the last figure charges, and he shoves a cart down the icy aisle into it. > **[23-24s]** SLOW MOTION: one final swing, and the last figure collapses and goes still. > **[24-26s]** Wide shot: the aisle falls silent, frost melting off the pole, lights flickering overhead. > **[26-28s]** He steps over scattered products, checking each aisle, chest heaving. > **[28-30s]** Final shot: he opens the stockroom door and the cashier steps out, relieved, as sirens approach outside. Cut to black. > > Natural sound: shelves and cans crashing, shouting, carts rattling, glass doors, heavy breathing, distant sirens, layered with a tense score that builds through the chaos and drops out for the final swing. Varied camera work: low angles, overhead security-cam shots, whip pans, tracking, handheld close-ups, with slow motion only on the last two takedowns.

Create a 30-second ultra-realistic cinematic sci-fi horror-thriller short film set inside a high-security underground biological research facility. The sequence begins with an immediate, violent hook at full speed. No slow motion until the designated moments. Maximum dynamic camera variety, grounded practical-effects filmmaking, realistic human reactions, and physical creature interaction. The horror must feel tangible, disturbing, and believable, like a premium theatrical sci-fi horror film. SETTING A massive underground research laboratory with reinforced glass chambers, steel walkways, industrial machinery, specimen containers, computer terminals, cables, and flickering fluorescent lights. Four scientists in protective laboratory clothing work around a sealed experimental chamber. A strange, dark, semi-transparent biological organism is suspended inside a cylindrical containment tank. Its surface shifts like living smoke trapped beneath liquid glass. Natural diegetic sound and an intense suspense score begin immediately. SEQUENCE — 30 SECONDS [0–1s] — THE BREACH Extreme close-up, real-time. A containment chamber's pressure gauge suddenly spikes into the red. The thick glass fractures with a violent explosive crack. Black translucent biological material bursts through the opening, spraying across nearby equipment. The camera jolts from the impact. [1–2s] — SOMETHING ALIVE Low-angle floor-level tracking shot racing alongside the spreading organism. The dark substance flows unnaturally against gravity, climbing a steel table leg while tendril-like structures snap outward at frightening speed. Electrical sparks flash in the background. [2–3s] — FIRST CONTACT Whip-pan to a scientist turning toward the noise. A thin black tendril shoots across the frame and wraps around his protective sleeve. His eyes widen. He attempts to pull away as the creature rapidly spreads across his arm. [3–4s] — THE TAKEOVER Overhead shot. The organism races across the scientist's body, forming a dark, translucent membrane over his protective clothing. He stumbles backward into a specimen table, knocking equipment onto the floor. Other scientists begin screaming and retreating. [4–5s] — PRACTICAL HORROR Extreme close-up of the scientist's gloved hand. The material pulses beneath the glove, the fabric stretches, and the fingers lock into a rigid, unnatural position. The creature pulls him violently out of frame. Avoid excessive graphic gore; emphasize physical transformation, sound, and terror. [5–6s] — PANIC Wide handheld shot pulling backward quickly. The remaining three scientists scatter through the laboratory. One knocks over a metal cart, another slams a security door, and the third grabs a portable industrial lamp. Red emergency lights activate. [6–7s] — IT DIVIDES Fast side-tracking shot. The organism splits into several narrow, rapidly moving tendrils that travel across the floor and up the walls. The camera follows one tendril as it disappears beneath a steel workstation. [7–8s] — SECOND ATTACK Low-angle shot from behind laboratory equipment. A tendril launches forward and catches a fleeing scientist around the ankle. He falls hard onto the floor, his ID badge skidding across the tiles. The camera shakes with the impact. [8–9s] — THE SCREAM Close-up of his face as the lights flicker. A dark membrane climbs toward his shoulder while he reaches desperately toward the others. The camera abruptly cuts away at the peak of his scream, leaving his fate unclear. [9–10s] — RUN Handheld chase-cam following the two remaining scientists sprinting through a narrow corridor between containment stations. Alarms blare. One glances over her shoulder as the organism crawls rapidly across the ceiling behind them. [10–11s] — THE COUNTERMEASURE Quick cut. The lead scientist tears open an emergency equipment cabinet and grabs a high-powered ultraviolet sterilization unit. Her hands shake as she activates it. The device emits a harsh electrical whine. [11–12s] — FIRST STRIKE Extreme close-up of the UV emitter igniting. She swings it toward the approaching organism and fires a concentrated ultraviolet beam. The beam illuminates the creature's translucent structure, revealing rapidly moving internal tendrils. [12–13s] — SLOW MOTION INSERT Slow motion reserved for this moment only. The organism recoils from the ultraviolet beam, its surface bubbling and contracting as its tendrils pull away from the light. Tiny particles burn away in the air. Make the practical effects tactile and realistic, not cartoonish. [13–14s] — THE TRAP Wide shot, real-time. The second surviving scientist rushes to a heavy industrial isolation gate and begins lowering it manually. The lead scientist maintains the UV beam while the creature surges toward them from multiple directions. [14–15s] — SURROUNDING THREAT Dynamic 360-degree tracking shot around the two scientists. The organism climbs the walls and stretches across the ceiling, attempting to bypass the gate. Their faces show exhaustion, fear, and intense concentration. [15–16s] — A RISKY DECISION Tight close-up on the lead scientist's face. She notices a pressurized cryogenic coolant line connected to the containment system. She looks at her partner and points toward it. A loud metallic impact comes from behind them. [16–17s] — BREAKING THE LINE Fast side-angle shot. The second scientist strikes the coolant valve with a heavy steel tool. The pipe ruptures, releasing a powerful cloud of freezing vapor across the corridor. Frost rapidly spreads over the nearby metal surfaces. [17–18s] — THE CREATURE FREEZES Overhead shot. The organism's tendrils stiffen as the freezing vapor surrounds it. Its movement becomes erratic and sluggish. The scientists retreat behind the partially lowered isolation gate. [18–19s] — FINAL SURGE Dynamic low-angle shot. The creature suddenly launches one last elongated tendril

Macro shot of rain on a neon-lit alley, camera pushes down into the puddle reflection and emerges in an alien desert at golden hour, ultra-detailed textures, volumetric light

FORMAT: 18 seconds, vertical 9:16. Raw authentic home video, indistinguishable from a real clip filmed by a family member on a cheap camera and posted online. Recorded on a mid-2000s consumer handycam in auto mode: small sensor, slightly soft lens, limited dynamic range, digital noise in the dark. One continuous handheld take, no cuts, real time. The camera and the operator never appear on screen: no reflections of them in the TV, window or any glass. REFERENCES: - IMAGE 1 (attached PS5 game page screenshot): defines ONLY the TV screen content after the game is selected. From 3.0s until the TV is hidden by the powder cloud, the TV shows IMAGE 1 exactly: the small square GTA VI game icon in the top-left corner; a white "PS5" badge followed by the title "Grand Theft Auto VI: Ultimate Edition" and a small circular loading icon; full-screen key art of a dark-haired woman with a high ponytail and hoop earrings embracing a short-haired stubbled man in a white tank top holding a pistol, against a pink-orange sunset sky, palm tree silhouettes, birds and a waterfront skyline; the pink-purple-orange "VI grand theft auto" logo on the left; the description "Experience the biggest, most immersive evolution of the Grand Theft Auto series."; the white rounded timer button with a clock icon reading "00:00:06"; the round grey "..." button; the dark "Friends Who Play" panel with a small avatar and "1 friend owns this game"; "Official news" at the bottom left; the bottom-right bar "R3 Health / Ratings / Privacy / Terms". Same layout, colors, fonts and positions. The ONLY change: the timer counts down from "00:00:06", same font, icon and position. The small top-left icon of IMAGE 1 is also the GTA VI tile on the PS5 home screen. IMAGE 1 does not affect people, room or camera. CAMERA BEHAVIOR: The operator is a family member standing behind the husband, filming him as a joke, holding the camera at chest height with one hand. Constant small hand tremor, breathing sway, horizon tilted a few degrees, framing slightly off, subjects drifting toward the edges. The operator is a few beats late on every event. Autofocus breathes and hunts: first on the bright TV, then late on faces; in the powder cloud it fails completely and the image goes soft. One clumsy digital zoom toward the TV that overshoots and pulls back. When the extinguisher fires, the operator flinches, turns half away, coughs, and the frame swings and dips before coming back. Auto-exposure reacts slowly: the TV overexposes the frame, the white cloud blows the image out to near-white, the dead TV plunges the room into grainy darkness and the exposure creeps back up over two seconds. The TV is filmed off the screen: visible moire, rolling scan banding, glare from the screen surface and slightly oversaturated colors. REALISM: Every detail must feel unstaged. Ordinary people, natural uneven faces, real skin with pores, under-eye shadows, stubble, oily forehead shine in the TV light, flyaway hairs. Clothes wrinkled and worn. Real reactions: delays, half-finished words, nervous laughter, squinting, coughing, spitting powder, wiping eyes with the back of the hand. Real room clutter: cables, an energy drink can, a plate with crumbs, a sock on the floor, chargers, dust on the stand. Real physics: the extinguisher is a heavy 6 kg red steel cylinder; she carries it low against her thigh, lifts it with visible effort, the discharge kicks back against her hands and the hose whips slightly. The discharge is real ABC dry chemical powder: a loud jet that turns into a dense pale yellow-white cloud filling the front of the room in a second, rolling slowly, reducing visibility to about one meter, hanging in the air, then sinking and coating every surface as fine matte dust. The TV is a real thin LCD: when hit, the panel cracks and bleeds black, the plastic bezel bends, the TV rocks on its feet. No movie exaggeration anywhere. CHARACTERS: - HUSBAND: around 30, average build with a slight belly, short messy dark hair, light stubble, faded grey hoodie with a small stain, dark jeans, mismatched socks. Cheap black gaming headset on his head. Kneeling on the rug close to the TV, gripping a white PS5 DualSense controller (white grips, black center, PS button between the sticks) with both hands, sweaty, trembling, completely absorbed. - WIFE: around 30, natural tired face, no makeup, long hair in a loose messy bun, faded pink fleece bathrobe over grey pajamas, worn house slippers. Calm, cold, unhurried, no shouting, the quiet of someone who has had enough. ENVIRONMENT: Small lived-in apartment living room at night, ceiling light off. A 55-inch thin LCD TV on two feet on a low cheap TV stand, a white PS5 (white side panels, black center body) standing vertically to the right of the TV, its light solid white. Rug, couch behind the camera, window on the right with cool city light and a slightly open curtain. The TV is the main light source: cold flickering light on his face and the walls, everything else dark. The doorway to the hallway is on the LEFT of frame, a faint warm light from the hallway. TIMELINE: 0.0-3.0s, SELECTING THE GAME: Camera behind his right shoulder. The TV shows the PS5 home screen: dark background, a horizontal row of square game tiles at the top. He presses right on the D-pad three times fast, the highlight moves tile by tile with soft ticks and stops on the GTA VI tile (same icon as the top-left of IMAGE 1), which enlarges slightly. He mutters: "There you are..." At 2.2s he presses X, soft confirm sound. The screen fades into the game page. 3.0-6.0s, COUNTDOWN: The TV shows IMAGE 1 exactly, timer "00:00:06". The clumsy zoom toward the screen overshoots and corrects; focus hunts on the key art, then sharpens on the timer. He shuffles closer on his knees, rocking: "Six... five..." Timer "00:00:05" at 4.0s. At 4.5s the wife walks in from the left doo

Create an ultra-realistic, cinematic action sequence featuring two young adult students, a blonde girl and a dark-haired boy, running across a beautiful urban beach. Scene 1 – The Chase: A young blonde woman wearing a stylish dark school uniform, white shirt, tie, blazer, and short pleated skirt runs energetically along the sandy beach. A young dark-haired man wearing a black school uniform runs beside her. Both look excited and determined. Their clothes and hair move naturally in the ocean breeze. Scene 2 – Dynamic Camera Movement: Capture fast-paced tracking shots, side angles, and close-up action shots as they sprint across the sand. Their footsteps kick up clouds of sand, creating a dramatic cinematic atmosphere. Scene 3 – The Giant Water Wave: Suddenly, an enormous magical wave rises behind them, forming a massive transparent water wall above the beach. The water sparkles beautifully under the sunlight, with realistic splashes, reflections, and flowing details. Scene 4 – Dramatic Ending: The two characters stop running and look back in shock at the gigantic wave. The boy falls onto the wet sand while the girl stands nearby, reacting dramatically. Show the powerful wave suspended in the background, creating a breathtaking cinematic moment. Visual Style: - Ultra-realistic 4K cinematic quality. - 16:9 widescreen composition. - Beautiful sunny beach environment. - Modern city skyline in the background. - Realistic ocean physics and water simulation. - Dynamic tracking camera movements. - Natural character movements and expressions. - Detailed sand particles, water splashes, and reflections. - Professional Hollywood-style action cinematography. - Smooth transitions and consistent character appearances. Important: Maintain the same two characters, outfits, hairstyles, and facial features throughout the entire video. Avoid distorted faces, unnatural body movements, and artificial-looking water effects.

[Overall Style and Core Story] Strictly 30 seconds, 16:9 horizontal screen, Seedance 2.5, live-action cinematic ancient-style xianxia twist comedy, native Mandarin audio-visual synchronization. Title: "What About the Group Photo?" The overall work adopts a genuine and natural Chinese ancient-style film and television aesthetic, with delicate skin texture, realistic silk fabric, natural volumetric light, rich environmental depth, and restrained character performance. It draws on information asymmetry, object usage, behavioral repetition, and role reversal from classic visual comedy, as well as the familiar real-life situation of constantly fussing over the people around you in order to take a satisfactory photo. The main storyline is very clear: The junior sister originally wanted to take a nice group photo together with the senior sister. But the senior sister in white is excessively concerned about whether she looks perfect in the frame. She keeps making the junior sister hold up a bronze mirror, adjust the direction, and find suitable lighting, gradually turning the junior sister into her exclusive lighting assistant. The audience realizes earlier than the junior sister that this so-called "group photo" actually only contains the senior sister. After the photo is taken, the junior sister discovers that only her hand holding the mirror appears in the frame. The senior sister even solemnly explains that "the hand was shot quite well." The junior sister does not argue, but instead hands the bronze mirror to the senior sister, letting her personally experience the process she just went through. The final shot fully echoes the opening, with only the two people's positions and identities swapped. The core comedy comes from: The overly serious senior sister; The junior sister who gradually senses something is wrong; The real-life contrast between one person pursuing the perfect shot and the other being forced to bear all the assistant work; And the precisely symmetrical role reversal at the end. No villain seeking revenge, no master undermining, no sudden combat, no random magic, or irrelevant twists are added. [Characters and Relationships] Character ID A: The same @Image 1 sword immortal senior sister. 25–30 year old adult East Asian female, maintaining exactly the same facial identity, bone structure, skin texture, proportions, temperament, and hairstyle as the reference image. Tall and slender, black long hair half tied up, secured with a white jade hairpin, always wearing a white embroidered silk hanfu, layered semi-transparent wide sleeves, a silver waist seal, a jade pendant, and white cloth boots, holding a single silver long sword. Calm, dignified, and confident in personality, usually appearing not to care about appearances, but actually quite particular about her own image. When it comes to having her photo taken, she arranges angles, poses, and lighting as seriously as if she were studying swordsmanship. Character ID B: The same @Image 2 junior sister. 20–25 year old adult East Asian female, maintaining exactly the same facial identity, features, facial proportions, braided hair, body type, and temperament as the reference image. Petite figure, naturally round and lively features, wearing the same set of teal linen hanfu, dark belt, wooden hairpin, and black cloth shoes, holding a single dark steel sword. Straightforward and observant in personality, initially willing to help the senior sister complete the photo shoot. When she realizes she has been treated as a mirror-holding assistant, she does not get angry or throw a fit, but simply lets the senior sister experience the same thing once. This episode features only these two core characters, with no additional master, enemies, or other major characters. The key props are an ancient polished bronze mirror about 30 centimeters in diameter and a xianxia image-recording stone stably placed on a stone platform. The bronze mirror relies on real natural light reflection for fill lighting, has actual weight, and is always reasonably held by the characters with both hands or one hand. The image-recording stone uses only a brief, soft spiritual glow to complete a single static image capture, with no modern cameras, electronic screens, or complex spell interfaces appearing. [Current Location and Spatial Continuity] All background and location reference images newly uploaded in this round jointly determine the actual physical world of the video. First silently extract and integrate the compatible real terrain, main architectural relationships, spatial scale, materials, vegetation, water bodies, weather, clouds, mountain mist, main light direction, colors, reflections, and atmospheric depth from the reference images. Recombine them into a unique, reasonable, and continuous actual scene for this round. Within these real spaces, choose a viewing platform or courtyard stone terrace suitable for taking a photo and safe to stand on; if the reference images do not have a corresponding platform, use a location that genuinely exists in the reference world and can accommodate character activity. Absolutely do not replace it with an unfamiliar generalized xianxia scene out of thin air. The entire story takes place in the same location. The image-recording stone is fixed on a stable stone base, facing the position where the senior sister originally stands. The senior sister initially stands at the center of the frame. The junior sister stands at the edge of the frame, holding up the bronze mirror to softly reflect natural light onto the senior sister's face. Most of the junior sister's body is outside the image-recording stone's framing range, with only her hand holding the bronze mirror and the edge of the mirror just barely entering the frame. This realistic geometric positioning must be maintained from the beginning all the way to the twist reveal, and must not suddenly change for the sake of a laugh. The background always remains naturally alive: Wind moves hair and sleeves; Leaves sway gently with the breeze; Existing water bodies flow continuously; Mountain mist changes slowly according to real terrain; Light and shadows and reflections change reasonably with the camera angle. But these environmental movements remain narratively neutral, not actively driving or changing the plot. [0–4 seconds: Opening hook] The opening first presents a very beautiful cinematic medium-long shot composition. The same senior sister in white stands in the magnificent natural space formed by the reference images. She poses elegantly, the silver long sword at her waist hanging naturally, her light-colored silk wide sleeves gently blown by the breeze. At first glance it looks like a truly gorgeous, solemn cinematic still of a sword immortal. However, the camera immediately performs a small, precise lateral movement. Beside the frame, the same junior sister in teal suddenly appears. The junior sister is laboriously holding up a bronze mirror, her arm already starting to ache. She looks toward the senior sister: Junior Sister: "Senior Sister, what about the group photo we agreed on?" The senior sister does not even turn her head, only looking at the position of the image-recording stone: Senior Sister: "Hold the mirror a bit higher." The originally solemn sword immortal image instantly turns into a all-too-familiar scene of photo-taking fussing. [4–10 seconds: Character goals and setup] Maintain the same location, the same bronze mirror, and the realistic spatial relationship between the two. Medium two-shot. The junior sister raises the bronze mirror a bit higher. The senior sister seriously observes the soft light reflected onto her face, gently moves her chin, and adjusts a strand of hair. Senior Sister: "A bit to the left." The junior sister moves her arm as requested. Senior Sister: "A bit higher." The junior sister complies again. The mirror has actual weight, and her raised arm is already visibly fatigued. The junior sister finally becomes somewhat puzzled: Junior Sister: "Senior Sister, am I in the frame too?" The senior sister glances at the framing direction of the image-recording stone. She has already noticed that most of the junior sister's body is not in the frame. A brief pause. The senior sister remains very calm: Senior Sister: "You're in, don't worry." The junior sister temporarily chooses to believe her. That extremely subtle unnaturalness on the senior sister's face makes the audience begin to realize that she is actually fully aware. [10–18 seconds: Comedy conflict escalation] Still maintain exactly the same location, characters, and prop relationships. The senior sister first slightly turns her body to find better light for her face, then adjusts the angle of the silver long sword to make her overall silhouette more elegant. She only changes her pose twice. But each adjustment requires the junior sister to readjust the bronze mirror angle to keep the reflected light stable. The camera does not use fast montage, but fully records two continuous, reasonable actions. While holding up the bronze mirror, the junior sister tries to stick her head into the image-recording stone's framing range, hoping she can also appear in the group photo. The senior sister immediately gently reminds her: Senior Sister: "Don't wobble, it's almost done." The junior sister has no choice but to stop. She maintains an increasingly uncomfortable mirror-holding posture. At this moment, give an objective shot from the actual framing direction of the image-recording stone. The audience can clearly see: At the center of the frame is the perfectly posed senior sister in white. At the edge of the frame only the junior sister's hand holding the bronze mirror and a bit of the mirror's frame are visible. The junior sister's face and body are completely outside the frame. This shot must let the audience understand the final punchline before the characters do. The image-recording stone emits only one extremely soft, brief spiritual glow to complete the static image capture. No exaggerated magic is produced. [18–25 seconds: Key change and twist] The image capture ends. The senior sister finally relaxes her pose: Senior Sister: "Done." The junior sister immediately lowers her already sore arm and walks over to the image-recording stone to check the image just taken. The camera first maintains a natural two-person medium shot, then briefly shows the composition just recorded. The captured result is completely consistent with the earlier framing shot: The senior sister stands beautifully at the center. Only one of the junior sister's hands holding the mirror appears at the edge of the frame. No text appears. The junior sister takes a serious look. A half-beat pause. Junior Sister: "Where am I?" The senior sister very seriously extends her hand and points to that hand at the edge of the frame. Senior Sister: "Here, the hand was shot quite well." The junior sister slowly turns her head to look at the senior sister. Junior Sister: "You call this a group photo?" The senior sister's originally confident expression shows obvious hesitation for the first time. She looks at the captured image, then at the junior sister. Finally realizing that her explanation really does sound somewhat outrageous. Do not let her continue to argue, and do not add a third person speaking. The humor in this segment relies on clear visual evidence, the senior sister's deadpan answer, and the junior sister's expression. [25–30 seconds: Punchline payoff and ending] The junior sister does not continue arguing. She simply very naturally hands the bronze mirror to the senior sister. Junior Sister: "Fine, your turn to hold it." The senior sister instinctively takes the bronze mirror. The junior sister has already walked to the center of the frame where the senior sister was just standing. She gently adjusts her teal sleeves and strikes a naturally relaxed, genuinely attractive ancient-style pose. The senior sister can now only stand at the edge of the frame, holding up the bronze mirror just as the junior sister did before. The camera returns to a composition almost identical to the opening. But this time a completely symmetrical role reversal has occurred: At the center is the beautiful junior sister in teal. At the edge is the senior sister in white holding up the bronze mirror. The junior sister looks toward the senior sister and says in an almost identical calm tone: Junior Sister: "Hold the mirror a bit higher." The senior sister looks at the junior sister. Then looks at the bronze mirror she is holding. She finally understands that the junior sister intends to give her exactly the same treatment she just received. The senior sister gently sighs. She can only resignedly raise the bronze mirror a bit higher. End precisely on this cinematic image that strictly echoes the opening, with the two characters' positions completely exchanged. No additional master's finishing blow, no added explanation, no extra sudden events. [Actor performance and comedy rhythm] The senior sister must treat this small matter seriously from beginning to end. Her humor comes from someone who clearly cares about her image but always presents herself with dignity and composure. The junior sister's emotions must develop continuously from trust and cooperation, to doubt and discovering the truth, and finally to calm counterattack. Arm fatigue gradually intensifies, but no exaggerated trembling, falling, or cartoonish movements are used. Key pauses include: The extremely brief hesitation before the senior sister answers "You're in, don't worry"; The half-beat silence of the junior sister after the captured image is revealed; The awkwardness after the senior sister says "the hand was shot quite well"; The helplessness when the senior sister is finally forced to hold up the mirror. All reactions are natural and believable. [Camera and editing] The opening uses a beautiful cinematic medium-long shot, then reveals the mirror-holding junior sister through a small lateral movement, allowing the audience to quickly reinterpret the image. The setup uses a two-person medium shot to clearly establish the spatial relationship. The escalation section shows actual actions through appropriately measured detail shots of the bronze mirror, arms, sword blade, and facial reflections, but does not use continuous rapid cuts. The image-recording stone's framing shot serves as core visual information and must clearly show who is in the frame and who is outside it. The twist is completed through the characters checking the captured result and their expression changes. The ending adopts a symmetrical composition that echoes the opening, allowing the audience to understand the role reversal without additional explanation. Camera changes are natural, with no showy orbiting and no arbitrary environment switching. [Mandarin dialogue and sound design] All dialogue uses natural Mandarin. The actual character lines remain completely consistent with the English portion. The senior sister's voice is clear and composed, with normal speaking pace, not deliberately aloof. The junior sister's voice is lively, cooperative in the first half, gradually showing confusion and dissatisfaction in the second half, and finally returning to calm with a touch of small satisfaction. Only one character is allowed to speak at a time. The speaking character's lip movements are accurate; non-speaking characters keep their mouths closed and respond with genuine eye contact and micro-expressions. Background sound consists of natural wind, distant flowing water, fabric rustling, soft clinks of the bronze mirror, character footsteps, and one faint spiritual glow sound from the image-recording stone. The music is extremely light, appropriately receding at the visual reveal and the final role reversal. No canned laughter, no on-screen subtitles or readable text. [Character, prop, and environment continuity] The senior sister and junior sister always maintain the facial identities, hairstyles, body types, clothing, and weapons corresponding to Image 1 and Image 2. Throughout the entire film there is only one bronze mirror and one image-recording stone. The bronze mirror's actual position, mirror surface direction, character grip method, and reflected light must all match each other. The junior sister's mirror-holding movements must conform to normal human range of motion. The senior sister's silver sword and the junior sister's dark steel sword must not be swapped. The image captured by the image-recording stone must be consistent with the actual framing composition shown earlier. The same location, terrain, architecture, trees, water bodies, main light direction, weather, and spatial layers remain continuous. [Seedance 2.5 execution requirements] Strictly 30 seconds, 16:9 horizontal screen. Prioritize generating the complete story in one pass, organized into five segments: 0–4 seconds, 4–10 seconds, 10–18 seconds, 18–25 seconds, and 25–30 seconds. Fully utilize Seedance 2.5's multi-character motion consistency, multi-reference material control, continuous shot narrative, and native audio-video generation capabilities. Camera shots can switch reasonably according to story needs, but all actions and prop states must be continuous. Do not add extra characters, dialogue, special effects, plot, or new locations. The real punchline must be clearly presented in the image. End with the character position swap completing the visual payoff, forming both a complete ending and preserving room for the two female leads to continue developing relationship comedy in the future. [Negative Prompt] blurry, bad quality, low quality, low resolution, noisy, jpeg artifacts, watermark, text, error; deformed, mutated, bad anatomy, poorly drawn hands, bad composition, out of frame, disfigured; inconsistent character, changing clothes, face morphing, background shift, glitching cuts, disappearing props

A beautiful young anime-inspired Korean girl with long silver-grey hair, delicate facial features, sparkling eyes, and star-shaped earrings, wearing a stylish burgundy cardigan over a patterned top, standing on a busy Japanese street at night. She is holding a freshly steamed white fluffy bao bun with both hands. She gently blows on the hot bun to cool it down, looks at it with excitement, then takes a small delicious bite and slowly chews with a cute, satisfied expression. Cinematic close-up shots of her face and the steaming bun, followed by a medium shot showing the beautiful city street, glowing shop lights, warm bokeh, and pedestrians in the background. Realistic food textures, natural facial expressions, subtle hair movement, soft warm cinematic lighting, shallow depth of field, ultra-detailed skin, smooth natural movements, premium anime-realism aesthetic, 4K quality. Vertical 9:16 video, 15 seconds, smooth cinematic camera movements, consistent character appearance throughout.

Create a 30-second, 1080p ultra-realistic cinematic commercial for a fictional premium travel luggage brand called VELA. A stylish young Korean woman with long, sleek dark hair and natural glowing makeup travels through a modern international train station carrying one premium VELA hard-shell suitcase. She wears a bold contemporary fashion look: a structured deep burgundy leather jacket worn over a fitted black asymmetric top, high-waisted wide-leg charcoal trousers with a dramatic tailored silhouette, pointed black ankle boots, sculptural silver earrings and a sleek black shoulder bag. The styling feels like a high-fashion Korean editorial while remaining realistic and wearable. Maintain the exact same woman, outfit and suitcase design throughout. The commercial opens with a low cinematic tracking shot of the VELA suitcase rolling smoothly beside her as she walks confidently through the busy station. Her bold outfit immediately establishes a strong fashion presence. The station feels completely real—natural travelers, luggage, footsteps, announcements, polished floors and authentic ambient movement. She changes direction through a crowded passage, effortlessly pulling the suitcase beside her. Show a detailed close-up of the quiet multi-directional wheels moving smoothly across the floor. She stops briefly near a seating area, places the suitcase upright and opens it. Inside, her clothing is neatly organized. She removes a folded piece of clothing, closes the suitcase and continues walking. She approaches an escalator and pulls the suitcase naturally behind her. Cut to a low tracking shot following the wheels as she moves toward a waiting modern train. She reaches the train, extends the suitcase handle, lifts the luggage naturally into the storage area and secures it. Show refined macro shots of the wheels, handle mechanism, textured shell, zipper and premium hardware. Inside the train, she sits beside the window with the VELA suitcase beside her. The train begins moving through the city. She looks toward the passing skyline, then down at her suitcase with a subtle confident smile. End with a cinematic hero shot of the same VELA suitcase, with the woman softly out of focus behind it. Voiceover “Travel doesn't always go as planned. That's why we designed for everything in between. Smooth when you move. Strong when you need it. Made to keep up with you. VELA. Ready for the journey.” Style: Ultra-photorealistic premium fashion-travel campaign, cinematic live-action, realistic Korean woman, bold contemporary styling, natural skin and hair, authentic human movement, physically accurate luggage physics, premium materials, realistic reflections, sophisticated lighting, smooth tracking shots, shallow depth of field, subtle film grain, luxury editorial cinematography. Negative: No outfit changes, no identity drift, no suitcase morphing, no changing luggage design, no duplicated people or luggage, no impossible physics, no floating objects, no CGI-looking environment, no futuristic effects, no holograms, no excessive slow motion, no distorted anatomy, no subtitles, no captions, no random text, no watermark.

Cinematic 3D animation, a fierce female martial arts student wearing a school uniform, gray blazer, pleated skirt, and white knee-high socks, single-handedly defeating multiple male opponents in a school hallway corridor. She executes powerful high kicks, swift punches, and throws opponents into lockers and onto the floor with dust and debris flying. Dynamic camera movements, intense slow-motion impact frames, dramatic high-contrast cinematic lighting with light streaming through hallway windows, photorealistic Unreal Engine 5 aesthetic, 9:16 vertical format

Cinematic epic fantasy battle: A high-intensity fight between two elemental warriors in a vast green valley surrounded by snow-capped mountains and a reflective lake under dramatic golden-hour clouds with volumetric god rays. Characters: - Young man with spiky white/silver hair, glowing bright blue eyes, dark hooded sleeveless jacket and pants. Body crackling with vivid blue lightning energy on arms, legs and fists. Agile martial-arts fighter. - Muscular dark-skinned warrior with long flowing black hair, glowing red-orange eyes, lava-like cracks on body, dark baggy pants with red sash. Wields a glowing orange-red katana and summons a massive fire dragon. Action sequence: White-haired warrior stands in shallow water as a massive orange fire blast erupts. He charges with blue energy trails. Close-up of his intense face. The two clash violently over water and grass — high kicks with blue lightning meeting the fire warrior’s sword. Rocks explode. Blue energy streaks across the field. Fire warrior summons a colossal serpentine fire dragon of roaring orange flames. Final confrontation: both fighters face off in shallow water with the massive fire dragon looming behind, blue lightning vs red fire energy clashing, water splashing, intense combat, sparks and explosions. Aerial wide shots of the vast valley. Style: Photorealistic, ultra-detailed, cinematic 4K, dramatic lighting, motion blur, volumetric lighting, particle effects for fire, lightning, water, dust and explosions, smooth dynamic camera movements including wide shots, close-ups, tracking and aerial views.

A cinematic macro product advertisement featuring a miniature pastel mint-green retro electric kettle inspired by vintage SMEG design. Two realistic human hands carefully unpack the tiny kettle from a small cardboard box and gently place it on a warm wooden kitchen countertop. Close-up shots show the hands pressing the kettle's lid button, opening the lid, and pouring crystal-clear water inside from a small glass pitcher. The kettle is switched on, a beautiful blue indicator light illuminates, and delicate realistic steam rises from the spout. Cozy vintage kitchen background with wooden shelves, ceramic jars, warm natural sunlight, shallow depth of field, soft bokeh, realistic reflections, highly detailed textures, smooth cinematic camera movements, premium luxury product commercial, photorealistic, 4K quality, vertical 9:16.

FORMAT 30-second HORIZONTAL 16:9 photorealistic cinematic zombie action film. One adult superpowered woman escapes through an unfinished skyscraper, crosses a steel beam bridge, descends a construction chute, escapes its collapsing outlet, and reaches an excavator. After sweeping nearby zombies aside, she discovers tactical officers aiming at her. PACE Urgent, sustained athletic sprinting at realistic human running speed: quick strides, strong foot plants, forward lean, driving arms, and hard breathing. Zombies sprint just behind her. Superstrength increases force, grip, and impact tolerance, NOT playback speed. No fast-forward, slow motion, speed ramps, floating jumps, or accelerated limbs. No waiting poses, walking, prolonged crouches, or reaction pauses. Landings flow into the next movement. Necessary balance adjustments remain brief and visible. Thirty seconds is EDITED screen time. Cuts omit travel, not decisive contacts or landings. Never replay a completed action. REFERENCE shows ONE adult woman from multiple angles. Preserve her exact face, natural skin texture, athletic build, short tousled pink-and-cobalt-blue bob, and pale hair clips. Copy her pink and pale-lavender athletic top, striped details, pink pleated athletic skirt with pale inset panels, white crew socks, and chunky white, lavender, and pink sneakers. Matching athletic undershorts remain discreet beneath the skirt. She is 20 years old. Functional action framing. No reference collage, studio background, sunglasses, backpack, jacket, weapons, white shirt, neck bow, or costume changes. ABILITIES Superhuman strength, grip, and impact tolerance with visible leverage, balance, effort, recoil, and weight. No flight, energy effects, hovering, or unsupported direction changes. ENVIRONMENT Unfinished skyscraper in a dense city, approximately 25 floors above the site. Exposed concrete slabs, steel columns, reinforcement, stored materials, dust. Overcast daylight and light wind. Gray-green surroundings contrast with her pink outfit. Route: floor opening -> steel beam bridge -> receiving slab -> construction chute -> sand mound -> excavator. The opening gap is 1.5 meters across, with a broad stable landing slab. The beam bridge lies only two running strides beyond this landing. The anchored I-beam spans a separate four-meter opening between two slabs. Its upper flange is approximately 40 centimeters wide. Both ends are firmly attached. It is a crossing, not a dead end. At the far end, one zombie approaches across the broad receiving slab. The heroine clears the beam before hitting it. A heavy rectangular steel formwork panel stands almost upright in a storage rack beside the exit. Its lower edge rests on concrete. One loose retaining clip holds its upper edge. Its fall path lies entirely on the receiving slab, across the beam exit, clear of her onward route. The chute entrance is three meters beyond the panel. Establish panel, beam exit, and chute together. CHUTE A wide open-top U-shaped steel trough with smooth worn metal, rounded joints, raised sides, and visible supports. Wide enough for an adult to slide feet-first with elbows tucked. It follows a substantial sloping construction ramp along the facade, then broad descending curves over an adjacent unfinished parking structure toward the site. Long slopes and broad turns; no vertical pipe, stairs, sharp corners, or unsupported sections. High and middle sections remain secure. Only the final three-meter section has a damaged outer mounting bracket. It sits less than two meters above the site and ends just above a deep sand mound. Its upper connection remains attached briefly when the bracket fails, so the outlet pivots downward toward the sand. No gap to jump. The section sags under her, then drops behind her after she exits. Establish the sand mound and excavator beyond it before her landing. SITE Yellow tracked excavator, sand mound, pipes, crane base, overturned mixer, chain-link fence. The excavator is a few strides beyond the sand, along her direction of travel. Engine already idling, cab door open, bucket low above the ground. Tracks remain stationary. The upper structure initially faces diagonally toward the road. One short rotation brings the cab toward the road while sweeping the bucket through zombies in front of the machine. Keep that sweep clear of the cab, tracks, sand mound, and fence. Zombies occupy the site inside the fence. Tactical officers stand on the road outside it. ZOMBIES Aggressive runners with pale gray skin, cloudy eyes, torn dirty construction clothes, occasional helmets and safety vests. No exposed organs or graphic dismemberment. Six to eight initially pursue her. Others occupy the receiving slab and site. Natural running, climbing, handholds, footholds, and weight. Enemies move independently without waiting their turn. Threat remains close. Her actions create brief advantages; enemies never freeze to give her time. TIMELINE 00:00-00:03 | OPENING JUMP Open on her final running foot plant before the 1.5-meter gap. A fast lateral tracking shot shows her full body, the landing slab, and zombies only two strides behind, with the city beyond the exposed floor. She pushes off, clears the gap in a compact forward leap, and lands with flexed knees. Her trailing foot immediately steps into the next running stride. Show takeoff and landing; no standing reset or roll. The leading zombie jumps after her and lands heavily but continues pursuing. Another stumbles at the edge. She takes the next two strides toward the beam already visible ahead. 00:03-00:06 | BEAM CROSSING Cut on forward motion to a three-quarter view establishing both ends of the beam and the receiving slab. She runs across with quick accurate foot placements and small arm adjustments

>> → OPENING FRAME. Clip starts exactly on this frame: woman with long black wavy hair, cream cap-sleeve blouse, olive wide-leg trousers, back to camera, walking into a grey concrete-block flower café drowned in sunflowers, single brass pendant lamp hanging centre, framed Van Gogh-style sunflower paintings on the wall, blurred books and camera in bottom foreground. >> also → CHARACTER REF — hair length and wave, build, posture, wardrobe. Lock likeness. Wardrobe never changes across the clip. >> also → LOCATION REF — concrete block walls, exposed timber beam, track spots, pendant lamp, trunk table with cake, matcha and phone, terracotta and dark glass vases, rusty filing cabinet, metal trunks, corner concrete pillar, mirror on the back wall. Every object keeps its real shape, scale and position in every shot. SECTION 1 — SHOT TIMELINE ONE CONTINUOUS ACTION, 14.7 seconds, 10 locked camera placements joined by hard match cuts. CAMERA RULE — every placement is a fully locked, tripod-dead frame. Zero shake, zero drift, zero zoom, zero push. The camera never moves inside a shot; it only changes position at the cut. Every cut lands mid-motion — her body, hands and hair pick up at the exact pose, speed and direction the previous frame left them. No resets, no repeated action, no pose jumps. Motion blur and depth of field consistent with a real cinema camera, 25fps, 1/50 shutter. PERFORMANCE: unhurried, no dialogue, no looking at the lens except the final frame. Natural weight shifts, hair swings with real inertia, fingers interact with petals, pages, ceramics with real contact and pressure. SHOT 1 (0.00–1.73) — ESTABLISHING, EYE LEVEL, LOCKED • Image 1 frame. She walks away from camera into the space, three slow steps, hair swinging, left hand brushing a sunflower head as she passes the trunk table. Pendant lamp top-centre, blurred books and camera body bottom foreground. Dust motes in the lamp pool. • CUT on her right foot lifting mid-stride, left hand still touching the petals. SHOT 2 (1.73–3.50) — FRONTAL WIDE, LOCKED, SYMMETRICAL • Camera now on the far side of the room facing back toward the entrance, centred on the tree trunk and the row of pendant lamps. Her right foot lands exactly where it was lifting; she is now walking toward camera, face half-lit, eyes down on the flowers, still brushing petals with her left hand as she passes. • CUT as her hand leaves the petals and her shoulders start to turn toward the back wall. SHOT 3 (3.50–4.77) — HIGH ANGLE MEDIUM, LOCKED, FROM TOP OF THE BACK SHELF • Camera mounted high, looking down past the mirror and sunflower painting. Her shoulders finish the same turn; she reaches with both arms, lifts the large burlap-wrapped bouquet off the floor against the wall. Only her back and side visible, face hidden by hair. Mirror catches a sliver of the pendant lamp. • CUT as the bouquet clears the floor and she begins lowering it toward the trunk table. SHOT 4 (4.77–6.00) — TRUE OVERHEAD FLAT LAY, LOCKED • Camera straight down on the trunk table: cake slice on an aged book, phone showing a music player, matcha latte with leaf art, sunflowers ringing the frame. The bouquet she was lowering enters top of frame and settles. Her right hand continues the same motion down, fingers wrap the warm cup and lift it out of frame. Brightest, most saturated shot. • CUT as the cup rises past frame edge, her fingertips the last thing visible. SHOT 5 (6.00–7.13) — HIGH ANGLE MEDIUM CLOSE, LOCKED, BOUQUET ABOVE STILL LIFE • Camera looking down at 45 degrees over the bouquet and table. The cup keeps rising in the same arc to her lips; she sips, steam thread rising, then sets the cup down on the trunk edge. Petals in sharp focus, concrete wall falling soft behind. • CUT as the cup touches wood and her other hand reaches for the open book. SHOT 6 (7.13–8.47) — HIGH ANGLE CLOSE, LOCKED, LOOKING DOWN AT HER LAP • She has sat on the low stool beside the trunk; camera is directly above her shoulder, looking down at olive trousers, open aged book, burlap bouquet across her lap. Her hand completes the reach, lifts the book from the table, and turns one page with a real paper flex. Nothing moves but page, hand and hair tips. • CUT as the page lands flat and she pushes up to stand. SHOT 7 (8.47–9.87) — STILL LIFE B-ROLL, EYE LEVEL, LOCKED, NO PEOPLE • Camera at shelf height on the sunflower field painting, dark glass vases, rusty filing cabinet, metal trunks. Pure hold. Her off-screen movement is implied only by a faint shadow crossing the cabinet door left to right and one vase petal trembling from her passing. • CUT exactly as the shadow leaves the cabinet. SHOT 8 (9.87–11.20) — LOW ANGLE, GRASS LEVEL, LOCKED, SHALLOW FOCUS • Camera on the floor among the sunflowers, lens almost touching petals, looking up. Blurred sunflowers fill the foreground, sharp sunflowers in the terracotta vase midground, ceiling beam and track spots far behind. Her olive trousers cross inches from the lens, heavy foreground blur sweeping through, then her arm reaches down into frame and pulls one stem from the vase, matching the standing-and-walking motion left by shot 7. • CUT as the stem comes free and her arm lifts. SHOT 9 (11.20–12.83) — PORTRAIT MEDIUM, EYE LEVEL, LOCKED • Camera across the room framing the corner, concrete pillar hard right. She arrives in the corner carrying the bouquet plus the single pulled stem, raising it in the same lift, and tilts her head slowly to the left, hair falling across her cheek. A bystander's head partly in frame bottom right, left in. Pendant light rims her hair. • CUT at the deepest point of the head tilt. SHOT 10 (12.83–14.53) — CLOSING WIDE, EYE LEVEL, LOCKED, SYMMETRICAL • Camera centred under the row of pendant lights, the same woman now completing the tilt back upright, standing

Duration: 30 seconds | Aspect: 16:9 | STYLE: Gritty, realistic action filmmaking with practical stunt physics, authentic film grain, natural motion blur, detailed textures, and imperfect cinema optics. No glossy CGI or video-game look. CHARACTERS Heroine: East Asian young woman with coral-orange hair, blunt bangs, white star-print bandana, thin metal glasses, black choker, white cropped T-shirt, navy polka-dot trousers, black cross-body bag, jetpack harness, and fingerless gloves. Enemy: Broad, muscular, olive-skinned man with dark stubble, scarred eyebrow, and matte-black tactical gear. Maintain identical faces, outfits, and proportions throughout. ENVIRONMENT AND LIGHTING Inside a grey turboprop cargo aircraft flying above the clouds. Industrial metal walls, cargo nets, containers, and a partially open rear ramp. Cold daylight enters through the ramp, creating dramatic silhouettes, deep shadows, and steel-blue lighting. Coral-orange hair, white clothing, and flying cash provide warm accents. 9-SHOT ACTION SEQUENCE SHOT 1 | 00:00–00:03 | ENTRY Wide establishing shot. The heroine enters the cargo hold through the open rear ramp as wind pushes loose banknotes across the metal floor. SHOT 2 | 00:03–00:06 | THEFT She opens a secured black case filled with cash, removes the security bands, and stuffs money into her cross-body bag. SHOT 3 | 00:06–00:09 | DISCOVERED The bands snap and individual banknotes explode into the air. A heavy boot appears behind her. The enemy reveals himself and lunges. SHOT 4 | 00:09–00:12 | FIRST ATTACK Handheld camera follows the enemy's punch. She ducks and redirects his arm, but he counters and slams her into a cargo net. SHOT 5 | 00:12–00:15 | FIGHT Fast whip-pans capture elbow strikes, a knee strike, and the enemy grabbing her clothing and forcing her backward. SHOT 6 | 00:15–00:19 | SLOW-MOTION COUNTER Low orbital camera movement. She catches his attacking arm, pivots, and delivers a powerful counter. Capture the impact at 120fps, played at 24fps, with flying money suspended around them. SHOT 7 | 00:19–00:22 | TAKEDOWN Return to real time. She uses his momentum to throw him onto the metal floor. Cargo straps bounce and banknotes scatter everywhere. SHOT 8 | 00:22–00:26 | FINAL STRIKE The enemy tries to rise. She delivers one final controlled strike, forcing him down. She looks toward the open ramp, breathing heavily. SHOT 9 | 00:26–00:30 | HERO SHOT Wide cinematic composition. She stands in the bright rear ramp light, holding her money-filled bag. The enemy lies defeated behind her. Banknotes slowly settle around her as she walks toward the open sky. Hold briefly, then hard cut to black. CINEMATOGRAPHY ARRI Alexa Mini LF, Alexa 65, Master Anamorphic lenses, and Laowa 24mm Probe Lens. Combine wide shots, close-ups, aggressive handheld movement, whip-pans, one orbital shot, low-angle impacts, and a slow final push-in. REALISM AND PHYSICS Authentic 35mm film grain, natural skin pores, sweat, fabric fibres, scratched metal, subtle lens imperfections, and realistic lighting. Money must consist of individual paper banknotes with visible fibres, folds, and creases. Every bill flutters, rotates, folds, and drifts independently in the aircraft's airflow. No rigid money blocks. All punches, falls, and collisions must have believable weight, recoil, and momentum. SOUND DESIGN Aircraft engine rumble, strong wind, fluttering paper, snapping currency bands, footsteps, heavy breathing, metallic impacts, and restrained cinematic music. No dialogue. NEGATIVE PROMPT No face changes, inconsistent clothing, duplicate characters, extra limbs, distorted hands, wire-fu, weightless movements, plastic skin, glossy CGI, video-game visuals, rigid money blocks, cartoon physics, excessive explosions, golden-hour lighting, text, subtitles, logos, or watermarks. FINAL REQUIREMENT: Exactly 30 seconds, 9 distinct shots, consistent characters, realistic stunt choreography, physically accurate flying paper currency, and gritty large-format photorealistic action cinema. FPS: 24fps, selected 120fps slow motion | Quality: 8K photorealistic cinema.

Create a 15-second ultra-realistic cinematic fashion-show video, 16:9 widescreen, 4K, 24fps, glamorous runway lighting, realistic crowd reactions, dynamic handheld camera movement, and high-end fashion-event atmosphere. Scene: A luxurious fashion runway packed with photographers, flashing cameras, and an excited audience. Two young, beautiful female models walk confidently toward each other on the runway, wearing elegant high-fashion outfits, heels, and professional runway makeup. 0–5 seconds: Both models confidently walk toward the camera, striking stylish poses as photographers take pictures. The first model stops and poses dramatically. 5–8 seconds: The second model also stops and poses. Suddenly, the first model becomes annoyed, turns toward her, and confronts her. Dialogue — Model 1: “Why are you giving a pose twice?” 8–12 seconds: The second model looks surprised and responds angrily. They argue briefly, then begin a dramatic but non-graphic fashion-runway scuffle, pushing each other lightly while trying not to ruin their outfits. 12–15 seconds: Security staff rush onto the runway and separate them. The audience reacts with shocked expressions while photographers continue capturing the hilarious unexpected moment. End with both models staring angrily at each other. Style: Photorealistic faces and skin, natural body movement, realistic lip-sync, authentic runway choreography, cinematic depth of field, dramatic spotlighting, detailed designer clothing, realistic crowd reactions, smooth motion, comedic timing, no injuries, no blood, no weapons.

Create a heartwarming, cinematic 3D animated cartoon video set inside a modern hospital room. A cute little baby wearing a white diaper and a soft white head bandage is resting on a hospital bed, with small bandages wrapped around the legs. The baby's worried mother and father stand beside the bed, looking concerned and emotional. Two friendly male doctors in blue medical scrubs and a cheerful female nurse wearing a blue uniform and white nurse cap carefully examine and treat the baby. The nurse gently approaches the baby with a medical syringe, while the baby looks curious and slightly nervous. As the nurse begins the treatment, cute pink heart-shaped graphics appear around the baby, creating a funny and heartwarming atmosphere. The baby gradually becomes happy and playful, smiling at the nurse. The parents feel relieved and smile with happiness. The doctors and nurse celebrate the baby's recovery together. Visual style: High-quality Pixar-inspired 3D animation, adorable expressive characters, large sparkling eyes, realistic soft skin textures, detailed hospital environment, cinematic lighting, smooth character movements, emotional facial expressions, beautiful depth of field, professional animated movie quality. Camera: Smooth cinematic camera movements, medium shots, close-ups of facial expressions, and wide shots showing the entire hospital room. Aspect ratio: 9:16 vertical, ultra-detailed, 4K quality, smooth animation, consistent character appearance throughout the video.

Foggy , muddy battlefield with a single bare tree in the middle. A lone knight in full spiked plate armour, chainmail and a dark red cape stands facing a barbarian chief.

Related models

Seedance 2.0

ByteDance's #1-ranked video model — multi-shot direction with native lip-sync and spatial audio.

Seedream 5

ByteDance's high-fidelity bilingual image model — the best quality-to-cost ratio for high-volume pipelines.

MiniMax H3

MiniMax H3 is a powerful multimodal video generation model that creates up to 15-second 2K videos with native stereo sound from text, image, video, and audio inputs.

Grok Imagine 2

xAI's Grok Imagine 2 — instant photorealistic stills that animate into short video clips with native audio, built into Grok.

Browse AI Video Prompts All models