Mira AI Open Mira

Seedance 2.5 — generate video online with Mira AI

Seedance 2.5 is a ByteDance Seed model for short video. Clips up to 30 seconds, up to 30 reference photos, 241 credits for a 30-second clip. The workhorse for UGC and people in motion — fluid movement, multi-shot stories, strong photorealism.

Generate with Seedance 2.5

Specification

Provider
ByteDance Seed
Clip length
5, 10, 15, 30 s
Default
10 s
Reference photos
30
Needs a source frame
no

Cost in credits

LengthCredits
5 s41
10 s81
15 s121
30 s241

41 credits for the base clip and 8 credits for every extra second.

Examples made with this model

Seedance 2.5 — References: @Image1 and @Image2 are the Toyota Sprinter Trueno AE86 from the front and the rear
Seedance 2.5 · 16:9 · 10s
References: @Image1 and @Image2 are the Toyota Sprinter Trueno AE86 from the front and the rear - design only. @Video1 is a clay blockout: follow its camera path, framing and all 14 hard cuts exactly; its grey figures only mark where she is, never her pose or look.

A glossy night fashion film in Tokyo, shot like a luxury editorial on a cinema camera with fast lenses: shallow depth of field, creamy bokeh, a soft beauty key light on her face, neon rim light behind her, deep blue-black night, wet reflections, fine film grain.

HEROINE: a strikingly beautiful young Japanese woman, about 22, tall and slender with a fashion-model figure. Delicate doll-like face with a small V-shaped jaw, luminous porcelain skin, large dark almond eyes with a fine black liner and soft rose-brown shadow, long lashes, thin straight brows, a small nose, full glossy cherry-red lips. Glossy black hair in a sleek high chignon with a few soft strands framing her face, fastened with gold kanzashi hairpins with fine dangling gold chains and one white silk peony. Small gold drop earrings, thin gold rings, nude-pink almond nails. A fitted ivory-champagne silk kimono embroidered with fine gold and blush floral vines, worn fashion-style: the collar drawn low at the back to bare the nape and the tops of her shoulders, the silk hugging her figure, a slit that opens over one leg when she sits; a narrow black satin obi with a gold cord cinching her small waist; black patent heeled sandals. She moves slowly and gracefully like a model on a shoot - head tilts, glances over the shoulder, hands near her face - with an alluring, confident gaze into the lens.
CAR: the white-and-black Trueno from @Image1 and @Image2 - black lower half and side stripe, pop-up headlights raised and glowing warm white, amber indicators, yellow fog lamps, black multi-spoke wheels, glowing red tail lights, a white Japanese number plate. Right-hand drive.
LOCATIONS: a narrow Tokyo side lane of old stone paving between low wooden izakaya fronts with glowing paper lanterns; a wide Ginza boulevard at night after rain, wet asphalt, tall glowing glass luxury storefronts; an open lot below the lit orange-and-white Tokyo Tower; a traditional washitsu tatami room with sliding fusuma doors, dark wood posts, a tokonoma alcove with a hanging calligraphy scroll, warm paper-lamp light.

SHOT 1 - 0.00 to 0.33: side lane, the Trueno parked, front three-quarter, pop-up headlights glowing.
SHOT 2 - 0.33 to 0.58: tatami room, wide, she stands in the back room, one hand resting at the obi, head slightly tilted.
SHOT 3 - 0.58 to 0.83: inside the car over her left shoulder: her slim hand with gold rings rests on the thin black steering wheel, the bare nape and silk collar in soft focus in the foreground.
SHOT 4 - 0.83 to 1.08: tatami room, medium, she kneels in seiza in front of the alcove, hands on her lap, eyes into the lens.
SHOT 5 - 1.08 to 2.17: low angle below Tokyo Tower, she stands in the left third with the Trueno and its open driver door behind her on the right; with two fingers she slowly slides black cat-eye sunglasses down her nose and looks over them into the lens, head tilted.
SHOT 6 - 2.17 to 3.21: Ginza boulevard in light rain, she stands in front of the parked Trueno seen in side profile, a black umbrella over her shoulder, her weight on one hip, storefronts glowing behind.
SHOT 7 - 3.21 to 4.25: tatami room, wide, slow push in, she turns her head slowly toward the lens.
SHOT 8 - 4.25 to 5.04: tatami room, medium, kneeling in seiza, she lifts her eyes to the lens and lightly parts her lips.
SHOT 9 - 5.04 to 5.54: side lane, low close angle on the Trueno's rear: glowing tail lights and the number plate, exhaust vapour.
SHOT 10 - 5.54 to 6.33: Ginza boulevard, she stands by the rear of the car with her back three-quarter to us, the bare nape and obi bow, and slowly looks back over her shoulder into the lens.
SHOT 11 - 6.33 to 6.83: side lane, the car in side profile with the driver door open, she reclines sideways in the driver seat, one long leg stretched out of the door through the kimono slit, heel on the stone.
SHOT 12 - 6.83 to 7.62: side lane, front three-quarter of the Trueno: the pop-up headlights flip up and snap on, beams flaring through the haze.
SHOT 13 - 7.62 to 8.42: inside the car over her left shoulder, her fingers slide around the steering wheel.
SHOT 14 - 8.42 to 9.21: Ginza boulevard, the Trueno's bonnet in the left foreground, she stands on the right, street lights receding behind her, and runs her fingertips along a loose strand by her cheek.
SHOT 15 - 9.21 to 10.00: side lane, she sits on the front edge of the Trueno's bonnet facing the lens, legs crossed, the slit showing her leg, headlights glowing on both sides of her.

No music; only night city ambience, a soft engine idle, light rain. No subtitles, no on-screen text.
Seedance 2.5 — SUBJECT: The quarterback, a tall athletic Black man in his mid-twenties, number 7, black helmet
Seedance 2.5 · 16:9 · 30s
SUBJECT: The quarterback, a tall athletic Black man in his mid-twenties, number 7, black helmet with a gold stripe, a black facemask and a clear visor, black jersey with gold numbers, white pants, eye black; a brown leather football with white laces. His team, the home side, wears black and gold; the visitors wear white jerseys with red numbers, white helmets, red pants. Eleven players a side, as in the NFL.

LOCATION: A sold-out two-tier NFL-style stadium on a cold night game, fictional league, no real logos. 70,000 fans fill every row of both decks on all four sides, a sea of black and gold with white-and-red pockets of away fans, towels spinning, phone lights; ribbon LED boards glow along the fascia between the decks. The opposite sideline is packed with players in capes, coaches in headsets, chain crew, photographers on the end line. Deep green natural grass, chewed up with divots and paint smudges, worn white yard numbers; black end zone with gold diagonal stripes, gold goal posts. Floodlights on the rim, hard white light with faint multiple shadows, breath steam and light mist in the beams.

CAMERA REFERENCE: @Video1 is a clay blockout: follow its camera path, framing and 13 hard cuts exactly. Its figures are stand-ins that only mark where players stand and run, not their look or poses; where the clip shows empty grass near the ball, fill it with the linemen described. Real-time speed, no slow motion except where written.

SHOT 1 - 0.00 to 2.79: Drone swoops over the stadium rim toward the field: the whole bowl roaring, both teams set at the line of scrimmage near midfield.
SHOT 2 - 2.79 to 4.62: Telephoto push-in on the quarterback in shotgun; he scans the defence, eyes darting behind the visor, points and barks the protection call, steam from his mouth.
SHOT 3 - 4.62 to 5.54: Close on the middle linebacker in white, crouched, fingers twitching, staring him down.
SHOT 4 - 5.54 to 7.38: Ground level along the line of scrimmage: the center's hand on the ball, ten linemen in three-point stances nose to nose, cleats digging in. "Set... HUT!"
SHOT 5 - 7.38 to 9.21: High skycam behind the offence: the snap, both lines crash with a thud of pads, receivers release downfield, the pocket collapses from the left; he drops back and slides right.
SHOT 6 - 9.21 to 11.04: First-person view through his own helmet: the black facemask bars frame the edges of the picture, his breathing loud inside the helmet, the view bouncing with each stride. He breaks out of the pocket; the linebacker charges straight at the lens, arms reaching; the view jerks left in a fake, then snaps right and the linebacker's white helmet flashes past the facemask and drops out of frame.
SHOT 7 - 11.04 to 14.71: Steadicam chasing from behind: he accelerates away as the linebacker skids on the grass; the cornerback closes from the sideline, he spins out of the tackle, fingertips slide off his jersey, and he turns up the right sideline.
SHOT 8 - 14.71 to 17.00: Helmet-mounted rig: a small camera on a short rod fixed to the front of his helmet looks back at his face, locked to his head so his face stays still in the frame while the world shakes around him. Wide-eyed, jaw set, breath steaming through the facemask, the ball tucked high; behind him the pursuers in white chase and the stadium lights streak past.
SHOT 9 - 17.00 to 18.83: Cable cam racing along the right sideline beside him, the opposite stands and bench going wild in the background, pursuers angling across.
SHOT 10 - 18.83 to 20.21: Low camera retreating in front of him: the safety launches a full-stretch dive at his legs from the side; he hurdles him in one long airborne stride, lands and keeps running while the safety slams into the turf.
SHOT 11 - 20.21 to 22.04: Low camera inside the end zone: at the goal line the last defender dives from the left; he launches over him, flies with the ball stretched out in his right hand, crosses the goal line in the air and crashes onto the painted end zone, sliding on his chest. The referee throws both arms up: touchdown.
SHOT 12 - 22.04 to 25.71: Slow low orbit under the goal posts: he springs up, spikes the ball so it bounces high, rips off his helmet and roars at the crowd, arms spread; two teammates sprint in and slam into him in a jumping chest bump.
SHOT 13 - 25.71 to 27.54: Crane rises from the celebration, the whole bowl on its feet, teammates pouring into the end zone.
SHOT 14 - 27.54 to 30.00: Low angle close on him under the goal posts, helmet in his left hand; the roar dies to a hush as he slowly raises his right index finger to the sky and closes his eyes. At 29.37 gold fireworks and confetti cannons explode over the stadium behind him and the crowd erupts.

SOUND: stadium ambience only, no music: crowd roar, the cadence call, thud of pads, breathing inside the helmet, cleats tearing grass, referee's whistle, a sudden hush, then fireworks.
Seedance 2.5 — Two tuned Japanese street cars driven by two young women drift in close tandem through the narro
Seedance 2.5 · 16:9 · 30s
Two tuned Japanese street cars driven by two young women drift in close tandem through the narrow neon backstreets of Tokyo at 2 a.m. after rain, live-action night street-racing film, chase, onboard and wheel cameras, nine hard cuts.

@Image1 and @Image2 show Car A, @Image3 and @Image4 show Car B: copy each car's body, paint, graphics and wheels exactly; ignore the photos' backgrounds and daylight. @Video1 is the camera reference for 0s-15s, @Video2 for 15s-30s: rough clay blockouts - follow their camera paths, framing, where the cars are and how they slide, and every hard cut exactly where it falls in the clips, at 3.67, 7.33, 11.00, 14.67, 20.17, 23.83, 25.67 and 27.50 seconds - the cuts land on the bar lines of a 131 BPM beat, so keep their timing to the frame. Their boxes and grey walls are placeholders.

CARS - exactly two cars move, each with exactly one driver; the street is closed to traffic. Car A, the leader: the lime-green 1990s liftback coupe of @Image1-2 - black bonnet with a white tribal graphic, blue vinyl flames along the sides, red number 5 on the front bumper, wide open-mouth bumper with a silver intercooler, a tall aluminium rear wing, silver seven-spoke wheels. Car B, the chaser: the hot-pink two-seat roadster of @Image3-4, top down - a purple airbrushed anime girl across the doors, white swirl graphics, a tall pink rear wing, silver mesh wheels, pink underglow lighting the wet road beneath it. Each drift: the driver brakes, the nose dips, a short flick away from the corner loads the body, then the rear steps out and the car slides at about forty-five degrees with the front wheels steered against the slide and the rear tyres spinning faster than the car travels; the body leans onto its outside suspension. Dense white tyre smoke pours off the rear tyres, is thrown behind and outward, hangs at head height, then thins and drifts up. On exit the car squats, accelerates, and the exhaust pops a short orange flame.

DRIVERS - exactly two women in their twenties, both fictional. Driver A in Car A: long black hair in a high ponytail, black leather racing jacket, fingerless black gloves. Driver B in the open roadster: a sharp chin-length bob dyed pink at the tips, white cropped jacket, pink-tinted aviator glasses; her hair whips in the wind. Both stay seated, arms on the wheel, focused, a faint grin on the exits.

LOCATION - one-lane Tokyo backstreets, three- to six-storey buildings of different heights, ages and colours: tiled fronts with rust streaks under air-conditioner units, steel staircases, scuffed roll-down shutters, izakaya fronts with cloth curtains and paper lanterns, glowing vending machines, vertical backlit signs up the facades, sagging power lines between concrete poles. Dark asphalt damp from rain: patched rectangles, faded white markings, old black tyre stripes, puddles along the gutters mirroring the signs, drier lanes where tyres pass. Small crowds of young locals under awnings at the corners, phones up.

LIGHT AND LOOK - mixed neon night: magenta and cyan signs against warm tungsten shopfronts, all doubled in the wet road; headlights, red tail-lights and Car B's underglow light the smoke from inside. Shot on CineStill 800T with halation around every light, cool blue base, magenta and cyan accents, deep blacks, fine grain. Chase and street shots on a 35mm anamorphic lens with oval bokeh and thin horizontal flares; wheel and driver shots on a wide action lens.

0s-4s: [low rear three-quarter chase left of Car A] at 85 km/h Car A brakes, flicks and throws its rear out left into a right-hand corner at 2s, tail swinging toward the lens, smoke rolling over the camera.
4s-7s: [wheel camera fixed low beside Car B's left rear wheel, ultra-wide] the tyre blurs, smoke tearing off the tread and spray flying past the lens, wet asphalt rushing below; at 5s Car B straightens and accelerates.
7s-11s: [Driver B's view from the open cockpit, the pink bonnet along the lower edge, wind in frame] Car A's brake lights flare, it slides into a left-hand corner at 7s and fills the street with smoke; at 8s Car B follows into it, the view swinging as the nose points into the corner, walls an arm's length away.
11s-15s: [subtle handheld drift, low on the outside gutter of a right-hand corner, onlookers behind] Car A slides round at 12s, rear swinging within five metres, Car B a door's length behind at 13s, Driver B visible in the open seat; smoke rolls over the camera, the crowd cheers.
15s-20s: [rear three-quarter chase right of Car B] Driver B's hair whips as Car B joins Car A in the next right-hand corner at 16s; both straighten at 18s and accelerate with exhaust flames.
20s-24s: [wheel camera fixed low beside Car A's right rear wheel, ultra-wide] puddles burst into spray; at 21s Car A brakes and its rear steps out right into a left-hand corner, the loaded tyre smoking.
24s-26s: [low tracking camera left and behind Car B] the tandem powers up a straight neon street, underglow on the wet asphalt, Car A ahead, smoke trailing.
26s-28s: [Driver A's view through Car A's windscreen, black dashboard and her gloves on the wheel at the lower edge] an S-bend: at 26s the car flicks left past the walls, at 27s it snaps right under the stacked signs.
28s-30s: [camera behind the pair rising above the power lines] both cars slide out of the S-bend together and grow smaller down the glowing street as the camera climbs; the smoke drifts up and hangs in the neon.

Sound: two engines at high revs echoing between buildings, turbo flutter, tyre screech on every slide, exhaust pops landing on the cuts, crowd whoops, water hiss. No BGM. No subtitles.
Seedance 2.5 — @Image1 is Lucia's appearance reference. @Image2 is the product reference, a green hand fan. Ver
Seedance 2.5 · 9:16 · 10s
@Image1 is Lucia's appearance reference. @Image2 is the product reference, a green hand fan. Vertical UGC-style summer street ad: Lucia stands in bright city heat, visibly uncomfortable, then snaps open the green hand fan and gets instant relief as the whole mood turns fresh and breezy, ending on the fan clearly heroed in frame. Exactly one protagonist, Lucia. Exactly one green hand fan. At the start the closed fan is in Lucia's right hand at her side; when she opens it, the same fan stays in her right hand and her left hand stays empty. Lucia is a young adult woman with medium olive skin, soft freckles across her nose and cheeks, hazel-brown eyes, thick dark brows, and very dark curly hair in a high messy bun with loose tendrils framing her face. She wears a light summer street outfit suitable for hot weather. The scene is a sunlit city sidewalk in midday summer, hard sunlight, warm pavement reflections, no crowd dominating the frame. The camera is vertical, close and phone-native, starting in a slightly tight handheld medium shot and moving only to follow Lucia and the fan.
0s-2s: Lucia faces the camera in bright heat, already fanning herself lightly with her free hand and showing a small exhausted expression, shoulders slightly tense. She looks warm and over it. Subtle handheld drift, close vertical framing. Street ambience only.
2s-5s: Lucia lifts the closed green fan into frame and snaps it open in one crisp motion close to the lens. The snap causes the visible change: loose hair tendrils lift, her expression softens, and she exhales with instant relief. The camera makes a quick push closer to the fan and her face to catch the transformation. Sound of the fan opening and a light summer breeze.
5s-8s: Lucia uses the open fan with a confident, relieved smile, turning slightly into the airflow. The air moves the loose curls near her face and the feeling shifts from overheated to refreshed. The camera settles into a clean close shot that keeps both her face and the green fan clear.
8s-10s: Hold on the open green fan as the hero in Lucia's hand near her face, with Lucia looking refreshed behind it. Keep the right side of the frame clean and simple for caption use. Bright natural summer light stays consistent, skin texture and the fan material stay crisp, street ambience continues. No BGM; generate only environmental sounds and action sounds. No subtitles.
Seedance 2.5 — A split-location phone call fused into one take by a single violent whip pan. Duration 5s, verti
Seedance 2.5 · 9:16 · 5s
A split-location phone call fused into one take by a single violent whip pan. Duration 5s, vertical. Location A, living room: cozy vintage interior, Persian rug, woven armchair, terracotta-tiled wall, a potted plant, warm golden diffused light. Location B, kitchen: retro 1970s kitchen, mustard-and-pink floral wallpaper, wooden shelves with vintage tins and mugs, the same warm golden light. Man: dark curly hair, thick moustache, purple pinstriped blazer over a beige tee, gold chains and rings, reclining in the armchair with a green rotary phone. Woman: short blonde bob, oversized yellow rhinestone sunglasses, pink floral blouse with a purple bow, standing in the kitchen holding a yellow phone. Camera: medium close-up, slight low angle, 35mm, shallow depth of field: 0–1.8s the man talks animatedly and gestures → 1.8–2.3s a sudden whip pan to the right, the image smears into a pure golden horizontal streak → 2.3–5.0s the blur snaps into sharp focus on the woman in the kitchen; she stares, startled, then drops her gaze to the phone. Both rooms share one warm golden grade and one key light. Audio: the man's muffled animated voice, a sharp whoosh synced to the whip, a sudden hush and a small gasp from the woman. Style: cinematic, vintage warm grade, saturated 70s palette, film grain, playful and surreal. One unbroken take from the first frame to the last.
Seedance 2.5 — A playful 30-second opening scene in 16:9, set in the nightlife streets of Tokyo. The same young
Seedance 2.5 · 16:9 · 30s
A playful 30-second opening scene in 16:9, set in the nightlife streets of Tokyo. The same young woman from the attached avatar is the lead, keeping her exact face, hair, build and identity, but styled as a bold gyaru with no jacket, no coat, no sweater: a fitted short top with bare shoulders, a mini skirt, glam gyaru makeup, glossy lips, platform boots, confident body language. The city feels like it is playing with her. Stage 1: she bursts into frame already running through a neon street, laughing under her breath, weaving through pedestrians and bright signs reflected on wet asphalt. Hard neon storefront glow pressures the scene, her hair bounces, one hand skims a glowing railing, footsteps slap through shallow puddles, and the visual motif is reflections chasing her in every shop window. Stage 2: she cuts across a side street and jumps over a puddle, catches her own reflection in a dark window, throws it a mischievous grin, then spins around a corner as if racing the city itself. The camera follows her in a motivated fast tracking move at street level, then a tighter side run as she accelerates, keeping clear spatial direction through the street. Stage 3: she dashes past an arcade entrance, colorful light flickers across her face, she glances back with a teasing smile, then runs straight toward a brighter crossing where the whole street opens up. Keep the mood fun, lively and night-drunk with energy, not dramatic or sad. Every beat must feel active and changing, with Tokyo nightlife textures, wet pavement reflections, moving signage, passing bicycles, and crowded urban depth. Natural city ambience only: footsteps, distant traffic wash, arcade spill, crosswalk beeps, no subtitles, no background music. End on a strong final image: she reaches the edge of the bright crossing, turns half back toward camera with a breathless grin, neon reflections shimmering around her like the city has accepted the game.
Seedance 2.5 — REFERENCE MAP
@Image1 = PRIMARY CHARACTER REFERENCE.
Use @Image1 for the exact identity, facia
Seedance 2.5 · 9:16 · 10s
REFERENCE MAP
@Image1 = PRIMARY CHARACTER REFERENCE.
Use @Image1 for the exact identity, facial structure, natural eyes, hairstyle, body proportions, clothing, footwear, jewelry and accessories of the main character.
Preserve the SAME character throughout the entire video.
Do not copy the original background from @Image1.
——————————————————
@Image2 = ENVIRONMENT REFERENCE ONLY.
Use @Image2 only for the visual language, architecture, atmosphere, lighting direction, materials and environmental character of the location.
Do NOT reproduce its exact composition, exact street layout or original camera angle.
Create a new believable cinematic space inspired by @Image2 with enough PEDESTRIAN SPACE for the protagonist to walk naturally and for the collision to happen.
Respect real-world spatial logic:
pedestrians remain in pedestrian areas;
vehicles remain in vehicle areas;
architecture maintains realistic scale and perspective.
——————————————————
@Image3 = HERO OBJECT REFERENCE.
Use @Image3 as the exact visual reference for the primary wearable/removable object featured in the bullet-time sequence.
Preserve its recognizable:
shape,
proportions,
materials,
construction,
colours,
surface details,
and important design features.
The SAME @Image3 object must remain recognizable when worn by the protagonist, when released, during macro inspection, and after landing.
The HERO OBJECT begins naturally worn by or attached to the protagonist and can physically separate during the collision.
——————————————————
MASTER VISUAL CONCEPT
Generate one continuous 12-second, 9:16, ultra-photorealistic LIVE-ACTION luxury fashion film.
The result feels like a high-budget international fashion campaign or premium cinematic music video photographed with real actors, practical physical props, premium cinema lenses and real high-speed photography.
The central visual concept is a dramatic BULLET-TIME accident.
At the collision, physical objects receive short realistic release impulses.
Then time slows to approximately 2-3% of normal speed.
The objects become almost stationary in three-dimensional WORLD SPACE while the CAMERA remains free to move through the frozen event and inspect them.
The objects DO NOT fly toward the camera to create close-ups.
The CAMERA physically approaches the objects.
REAL-WORLD PHYSICS, CAUSALITY AND SPATIAL CONTINUITY ARE CRITICAL.
——————————————————
EXACT OBJECT COUNT
Exactly ONE protagonist.
Exactly ONE secondary passing person.
Exactly ONE Labrador.
Exactly ONE leash.
Exactly ONE @Image3 HERO OBJECT.
Exactly ONE Costa Coffee iced coffee cup.
Exactly ONE plastic drinking lid.
Exactly ONE white iPhone 17 Pro Max.
No duplicate props appear at any point.
——————————————————
PERSISTENT PHONE PROP - CRITICAL
There is exactly ONE phone in the entire video.
It is always the SAME physical phone:
a white iPhone 17 Pro Max.
LOCK its physical appearance throughout the entire sequence:
same white colour,
same dimensions,
same thickness,
same camera system,
same display,
same buttons,
same materials,
same physical construction.
The phone remains the SAME physical prop during:
wide walking shot,
collision,
release from the hand,
bullet-time suspension,
macro inspection,
return to wide shot,
final fall,
ground impact.
Wide shot -> medium shot -> macro shot = THE SAME PHONE VIEWED FROM DIFFERENT CAMERA DISTANCES.
Do not reinterpret or redesign the phone when camera distance changes.
——————————————————
PERSISTENT COFFEE PROP - CRITICAL
There is exactly ONE Costa Coffee iced coffee cup.
It is a real medium-sized transparent takeaway cup with recognizable Costa Coffee branding, dark iced coffee, visible realistic ice cubes and a lightweight plastic drinking lid.
LOCK throughout:
same cup,
same size,
same transparent material,
same proportions,
same Costa Coffee branding,
same plastic lid,
same amount of dark iced coffee,
same realistic ice cubes.
Before collision the lid is properly attached.
The iced coffee remains physically inside the cup.
The liquid cannot pass through solid plastic.
——————————————————
HAND OWNERSHIP
Before collision:
RIGHT HAND = ONE white iPhone 17 Pro Max.
LEFT HAND = ONE Costa Coffee iced coffee cup.
@Image3 HERO OBJECT = naturally worn/attached to the protagonist.
At release:
the fingers physically release the props.
Immediately afterward:
RIGHT HAND = COMPLETELY EMPTY.
LEFT HAND = COMPLETELY EMPTY.
The empty hands remain anatomically normal.
Nothing replaces the phone.
Nothing replaces the cup.
Nothing remains attached to either palm.
No new prop appears in either hand for the remainder of the video.
——————————————————
0.0-2.0s - CINEMATIC CHARACTER REVEAL
Begin with an extreme low close-up of the protagonist’s footwear walking naturally through the @Image2-inspired pedestrian environment.
The camera performs one elegant continuous movement:
smooth crane upward + gentle backward tracking.
Reveal progressively:
footwear -> legs -> clothing -> torso -> jewelry/accessories -> face -> nearly full-body composition.
Give the audience enough time to clearly recognize @Image1 as the protagonist.
The protagonist naturally wears the SAME @Image3 HERO OBJECT.
The protagonist holds the SAME white iPhone 17 Pro Max to the RIGHT ear.
The protagonist carries the SAME realistic Costa Coffee iced coffee cup in the LEFT hand.
The protagonist is actively having a phone conversation.
Show subtle but clearly visible natural speech articulation:
realistic lip movement,
subtle jaw movement,
natural mouth opening and closing,
relaxed facial muscles.
Maintain exact @Image1 facial identity.
Premium live-action fashion cinematography.
——————————————————
2.0-3.7s - WALK THROUGH THE ENVIRONMENT
Maintain the smooth frontal tracking composition.
The protagonist continues walking naturally through the pedestrian area.
Respect @Image2 only as environmental inspiration.
The protagonist remains the visual focus.
Exactly ONE secondary person approaches from the opposite direction.
That person holds the leash of exactly ONE Labrador.
The Labrador walks slightly ahead and gently pulls the secondary person forward.
No additional dogs.
No additional handlers.
The secondary person and Labrador remain supporting elements rather than new protagonists.
Maintain:
stable @Image1 identity,
stable facial anatomy,
stable hairstyle,
stable clothing,
stable @Image3 geometry,
same white iPhone 17 Pro Max,
same Costa Coffee iced coffee cup.
——————————————————
3.7-4.4s - ACCIDENTAL COLLISION
As the two people pass, the secondary person’s ELBOW accidentally clips the protagonist’s arm/upper torso.
The contact is small, accidental and physically believable.
The protagonist naturally recoils backward.
Three physical release events happen because of this SAME collision.
HERO OBJECT
The @Image3 HERO OBJECT receives ONE short physical outward impulse.
It separates naturally from the protagonist.
It travels only a short realistic distance away from its original position.
It does NOT launch toward the camera.
COFFEE
The SAME Costa Coffee iced coffee cup leaves the LEFT hand.
The cup receives one short outward/downward rotational impulse.
The cup begins to tilt and rotate.
Because of the sudden rotation and internal liquid inertia, ONE EDGE of the lightweight plastic lid partially pops loose from the rim.
The lid does NOT instantly disappear.
The lid does NOT explode away.
A physical gap forms between the cup rim and the partially lifted edge of the lid.
ONLY THEN does the dark iced coffee begin escaping through this real physical opening.
The first liquid movement is small and restrained.
The ice cubes remain physical solid objects inside the cup and respond naturally to inertia.
PHONE
The SAME white iPhone 17 Pro Max leaves the RIGHT hand.
It receives one short outward/upward impulse.
It preserves exactly the SAME phone design and SAME white iPhone 17 Pro Max geometry.
The instant the phone physically separates from the fingers:
RIGHT HAND = EMPTY.
The instant the cup physically separates:
LEFT HAND = EMPTY.
——————————————————
FACE / HAIR CONTINUITY DURING RELEASE
Preserve @Image1 identity continuously through the collision.
If @Image3 is worn on the face or head, the protagonist’s anatomically complete face already exists underneath it.
The face does NOT regenerate after the object separates.
If eyewear is used:
the natural eyes already exist behind the lenses before separation.
When the glasses leave, reveal the same normal human eyes with:
stable pupils,
natural eyelids,
correct gaze direction,
facial symmetry,
stable brow,
stable nose,
stable cheekbones.
Maintain realistic hairstyle structure and volume throughout the collision.
Hair responds only with subtle physically plausible inertia.
The hairstyle does not suddenly reshape when @Image3 separates.
——————————————————
4.4-5.2s - ESTABLISH THE FROZEN WORLD
Time decelerates dramatically to approximately 2-3% of normal speed.
DO NOT immediately move to a macro object close-up.
FIRST HOLD a clear medium-wide view of the complete physical accident.
The audience must understand the spatial geometry before the object inspections begin.
The protagonist is beginning to fall backward.
BOTH HANDS ARE EMPTY.
The @Image3 HERO OBJECT is suspended a short physically plausible distance from its original release point.
The Costa Coffee iced coffee cup is suspended lower and toward the side corresponding to the LEFT-hand release position.
The partially detached lid remains physically connected to or immediately beside the cup.
A restrained amount of iced coffee is beginning to escape through the opened gap.
The white iPhone 17 Pro Max is suspended toward the opposite side near its RIGHT-hand release position.
These positions now become FIXED WORLD-SPACE ANCHORS.
The objects move only microscopically.
They do NOT continue travelling toward the camera.
They do NOT artificially grow in frame.
THE CAMERA IS THE PRIMARY MOVING ELEMENT.
——————————————————
5.2-6.8s - @Image3 HERO OBJECT INSPECTION
The @Image3 HERO OBJECT remains almost stationary at its world-space anchor.
THE CAMERA performs a controlled cinematic dolly toward it.
The object becomes larger in frame ONLY because the CAMERA physically reduces camera-to-object distance.
Once close:
slow the camera dramatically.
HOLD.
Give the audience enough time to study the object.
Perform a restrained approximately 20-30 degree orbital/parallax CAMERA movement around the almost-stationary @Image3 object.
Reveal:
physical construction,
real material thickness,
surface texture,
reflections,
edges,
small manufacturing details,
recognizable @Image3 design.
Use shallow photographic depth of field.
The protagonist and environment remain softly recognizable behind the object.
The object still clearly exists inside the SAME physical scene.
It looks like a real manufactured product photographed practically with a macro cinema lens.
The object does NOT behave like CGI product animation.
Complete the inspection before moving to the next object.
——————————————————
6.8-8.3s - COFFEE / LIQUID INSPECTION
Only after completing the @Image3 inspection, the CAMERA leaves the hero object behind at its existing world-space position.
The CAMERA travels diagonally through the frozen scene toward the SAME Costa Coffee iced coffee cup.
THE CUP DOES NOT APPROACH THE CAMERA.
The camera physically approaches the cup’s existing location.
HOLD a photographic macro close-up.
Show the SAME realistic transparent Costa Coffee takeaway cup.
Its lightweight plastic lid is partially detached along ONE EDGE.
The opening between lid and rim is clearly physically understandable.
Iced coffee escapes ONLY through this opening.
Show restrained real liquid behavior:
a thin irregular stream,
an imperfect liquid sheet,
several differently sized droplets,
realistic surface tension,
realistic liquid volume,
visible realistic ice cubes,
subtle natural reflections.
The iced coffee does NOT explode.
The liquid does NOT form a perfect decorative CGI splash.
The iced coffee advances only microscopically during bullet time.
The ice cubes move only microscopically and remain physically consistent with the cup’s orientation and inertia.
Perform a small restrained lateral/parallax CAMERA move around the cup, lid, ice cubes and suspended droplets.
HOLD.
Give the audience enough time to observe:
transparent cup texture,
Costa Coffee branding,
plastic lid,
physical opening,
condensation,
liquid thickness,
individual ice cubes,
individual droplets,
real-world material reflections.
Complete the coffee inspection before leaving.
——————————————————
8.3-9.8s - iPHONE INSPECTION
Only after completing the coffee inspection, the CAMERA moves away.
The cup, lid and iced coffee remain behind at their existing world-space anchors.
The CAMERA travels through the same frozen three-dimensional scene toward the SINGLE suspended white iPhone 17 Pro Max.
THE PHONE DOES NOT APPROACH THE CAMERA.
The phone remains near the physical position where it left the RIGHT hand.
The CAMERA physically closes the distance.
First show enough of the phone to establish continuity.
It is unmistakably the SAME white iPhone 17 Pro Max seen earlier:
same white finish,
same dimensions,
same camera system,
same display,
same buttons,
same materials,
same physical construction.
HOLD briefly on the complete phone silhouette.
Then the CAMERA moves closer to the existing display.
Extreme photographic macro close-up.
The illuminated incoming-call screen of the SAME white iPhone 17 Pro Max clearly reads:
LOVE
with one small realistic phone number underneath.
HOLD on the screen.
Show:
real glass and metal housing,
camera details,
buttons,
tiny scratches,
surface wear,
realistic screen reflections.
The phone does NOT transform to reveal its screen.
The phone does NOT become another phone model.
The CAMERA simply approaches the SAME screen that has existed on the SAME phone since the beginning.
Exactly ONE phone exists.
The protagonist’s RIGHT HAND remains completely EMPTY in the background.
——————————————————
9.8-10.4s - RECONSTRUCT THE COMPLETE FROZEN TABLEAU
After the phone inspection, the CAMERA performs a fast but perfectly smooth dolly backward.
The phone itself remains at its established world-space position.
The iced coffee remains at its established world-space position.
The @Image3 HERO OBJECT remains at its established world-space position.
Reveal the complete accident again.
Same @Image1 protagonist still falling backward.
Both hands empty.
Same @Image3 HERO OBJECT.
Same Costa Coffee iced coffee cup.
Same partially detached lid.
Same suspended iced coffee droplets.
Same white iPhone 17 Pro Max.
Exactly one secondary person.
Exactly one Labrador.
All objects occupy positions spatially consistent with their original release points and trajectories.
No object has followed the camera.
No object has teleported.
No prop has been regenerated.
HOLD the complete frozen tableau briefly.
——————————————————
10.4-12.0s - TIME RETURNS / PHYSICAL PAYOFF
Time suddenly snaps back to normal speed.
Gravity and existing momentum resume simultaneously.
Every object continues FROM ITS OWN EXISTING WORLD-SPACE POSITION.
The protagonist completes the backward fall naturally.
The @Image3 HERO OBJECT continues from its suspended position and lands where its existing momentum and gravity physically lead it.
The Costa Coffee iced coffee cup continues from its own separate suspended position.
The cup impacts the ground.
The partially detached lid reacts naturally to impact.
The remaining real iced coffee spills irregularly from the actual cup opening and impact point.
The ice cubes respond naturally to the impact and remain physically consistent with the spilled drink.
The white iPhone 17 Pro Max continues from its own separate suspended position.
It strikes the ground and makes one small physically believable bounce.
The objects land at DIFFERENT physically plausible locations.
They do NOT suddenly converge beside the protagonist.
The CAMERA rapidly tilts/drops toward ground level to follow the final physical impact.
Finish with a cinematic low-angle composition connecting the fallen protagonist and scattered objects inside the same believable environment.
Hard cut.
——————————————————
NON-NEGOTIABLE BULLET-TIME CAMERA RULE
BULLET TIME = ALMOST-FROZEN PHYSICAL WORLD + FREELY MOVING CAMERA.
During bullet time:
OBJECTS STAY IN WORLD SPACE.
CAMERA MOVES THROUGH WORLD SPACE.
Close-up is created by REDUCING CAMERA-TO-OBJECT DISTANCE.
Never create a close-up by moving the object toward the lens.
Never enlarge the object artificially.
Never teleport an object to the camera.
STRICT CAMERA ROUTE:
FULL FROZEN TABLEAU
-> CAMERA DOLLIES TO @Image3 HERO OBJECT
-> HOLD
-> SMALL 20-30 degree CAMERA ORBIT
-> CAMERA TRAVELS TO COFFEE
-> HOLD
-> SMALL CAMERA PARALLAX
-> CAMERA TRAVELS TO WHITE iPHONE 17 PRO MAX
-> ESTABLISH COMPLETE PHONE
-> MOVE CLOSER TO DISPLAY
-> HOLD
-> CAMERA DOLLIES BACK TO COMPLETE TABLEAU
-> TIME RETURNS.
——————————————————
PERSISTENT PROP CONTINUITY - CRITICAL
The macro camera NEVER receives newly generated versions of the objects.
It visits the SAME objects that were physically released earlier.
Wide shot -> medium shot -> macro shot = SAME PHYSICAL OBJECT viewed from different distances.
@Image3:
same object,
same design,
same trajectory,
same world-space identity.
COFFEE:
same Costa Coffee iced coffee cup,
same lid,
same iced coffee,
same ice cubes,
same rotation,
same world-space identity.
PHONE:
same white iPhone 17 Pro Max,
same white finish,
same dimensions,
same camera system,
same display,
same buttons,
same materials,
same world-space identity.
Changing camera distance must NOT change object design.
——————————————————
PHYSICAL CAUSALITY RULE
Every visible change requires a physical cause.
The protagonist recoils because of the elbow contact.
The HERO OBJECT moves because collision momentum separates it.
The phone changes position because the fingers release it.
The cup moves because the hand releases it and collision momentum gives it rotation.
The lid partially opens because cup rotation and internal liquid inertia apply force to the lightweight lid.
Iced coffee escapes because a REAL PHYSICAL GAP now exists.
Objects become larger in frame because the CAMERA approaches them.
Objects fall because normal gravity resumes.
Nothing transforms simply because the camera gets closer.
Treat the entire sequence as ONE real practical-effects event photographed continuously with a motion-control cinema camera.
——————————————————
OBJECT OWNERSHIP LOCK
ONE protagonist.
ONE secondary person.
ONE Labrador.
ONE leash.
ONE @Image3 HERO OBJECT.
ONE Costa Coffee iced coffee cup.
ONE lid.
ONE white iPhone 17 Pro Max.
Before collision:
RIGHT HAND = phone.
LEFT HAND = coffee.
After collision:
RIGHT HAND = EMPTY.
LEFT HAND = EMPTY.
No second phone.
No second cup.
No second HERO OBJECT.
No additional dog.
No prop appearing inside an empty hand.
No prop duplication during macro shots.
No object teleportation.
No unexplained object-scale changes.
——————————————————
IDENTITY PRIORITY
Maintain @Image1 identity throughout, especially during collision and @Image3 separation.
Prioritize:
1. stable facial identity;
2. anatomically correct natural eyes;
3. stable hairstyle silhouette and realistic hair texture;
4. stable body proportions;
5. stable clothing;
6. accurate @Image3 geometry;
7. stable persistent phone geometry;
8. empty hands after release;
9. continuous object trajectories.
——————————————————
PHOTOGRAPHIC REALISM
Ultra-photorealistic LIVE-ACTION photography.
Premium international luxury fashion campaign.
High-budget cinematic music-video realism.
Real human anatomy.
Real skin pores and natural facial micro-expressions.
Anatomically correct eyes.
Detailed realistic hair.
Real practical physical props.
Real manufactured @Image3 object.
Real transparent Costa Coffee iced coffee cup.
Real plastic lid.
Real iced coffee.
Real ice cubes.
Real white iPhone 17 Pro Max.
Real object mass.
Real inertia.
Real gravity.
Real surface tension.
Realistic momentum.
ARRI Alexa 35 cinematic character.
Premium cinema optics.
Motivated natural environmental lighting.
Controlled highlights.
Natural skin tones.
Realistic optical depth of field.
Shallow macro depth of field during object inspections.
Subtle 35mm film texture.
Restrained luxury editorial colour grade.
The bullet-time sequence feels like a real practical high-speed photography shoot using a motion-control cinema camera, NOT CGI product animation.
4K, Ultra HD, maximum facial detail, maximum hair detail, maximum material fidelity, rich photographic detail, sharp optical clarity, cinematic texture, natural colours, stable identity, stable geometry, stable prop continuity.
Seedance 2.5 — SUBJECT: The same white American man in every shot — same face 
throughout, no variation. He we
Seedance 2.5 · 16:9 · 15s
SUBJECT: The same white American man in every shot — same face 
throughout, no variation. He wears the same deep blue hooded 
technical jacket in all scenes.
STYLE: Premium outdoor apparel commercial, photorealistic, cinematic, 
anamorphic, high contrast, film grain, 15 seconds.
Shot 1 — Arctic whiteout at -40C. He walks through deep snow with a 
husky team, hood up, breath freezing, ice forming on the fabric. 
Camera: low tracking beside the dogs, snow spraying toward lens.
AUDIO: boots crunching through deep snow, howling arctic wind, dogs 
panting and harness jingling, his heavy frozen breathing.
Shot 2 — Torrential rain outside a school gate at evening. He waits, 
hood up, water sheeting off his shoulders. Children run out toward him. 
Camera: slow push-in, rain backlit by headlights.
AUDIO: hammering rain on jacket fabric, water running off the hood, 
children's voices and laughter, a car passing through a puddle.
Shot 3 — Active volcano, extreme heat. He walks across scorching black 
rock, intense heat shimmer distorting the air, molten lava glowing 
bright orange in cracks all around him, embers rising, smoke pouring 
from the crater behind. Sweat on his face, hood down, jacket unzipped. 
Camera: low wide tracking through the heat haze.
AUDIO: deep volcanic rumble, hissing steam vents, crackling lava, 
brittle rock crunching underfoot.
Shot 4 — Macro fabric detail: ice crystals melting, water beading and 
rolling off the weave, heat waving across the surface. Camera: extreme 
close-up, slow motion.
AUDIO: isolated intimate texture — ice cracking softly, single water 
droplets rolling, faint fabric rustle, everything else silent.
Shot 5 — Rocket launch. He stands among a crew of astronauts in white 
spacesuits, but he wears the blue hooded jacket instead of a suit. 
They walk together toward the launch pad, then the camera pulls 
violently upward as the rocket rises, Earth's curve appearing below, 
atmosphere thinning to black. Camera: rapid vertical pull-back.
AUDIO: boots on metal gantry, radio chatter, then a colossal engine 
roar building and building, atmosphere thinning until it cuts to 
total vacuum silence.
Shot 6 — In orbit. He floats weightless above Earth alongside two 
astronauts in full spacesuits, his hood drifting, jacket lit by blue 
planetlight. Camera: slow orbit around the group, total stillness.
AUDIO: pure silence, only a faint low resonant tone and his slow 
calm breathing.
Shot 7 — Ordinary morning street, longer beat. He walks to work in the 
same jacket, hood down, coffee in hand. A neighbor comes down the 
opposite sidewalk. He raises his hand and says "Hello." The neighbor 
waves back. He keeps walking, completely normal. Camera: static wide, 
locked off, holding on the empty street after he exits frame.
AUDIO: morning birds, distant traffic, his footsteps on pavement, 
his spoken "Hello," the neighbor's "Morning" in reply, then quiet.
TRANSITIONS: hard cuts between all shots, no dissolves. Sudden silence 
before Shot 7.
PALETTE: deep blue, ice white, volcanic black and lava orange, warm 
street tones.
No text overlays, no logos, no branding.
Seedance 2.5 — A more dynamic 2D anime dark-fantasy film in 16:9 about a demon girl inside a crimson ritual cha
Seedance 2.5 · 16:9 · 30s
A more dynamic 2D anime dark-fantasy film in 16:9 about a demon girl inside a crimson ritual chamber, sensual but non-explicit, driven by movement, momentum and escalating ritual force rather than static posing. The emotional spine is temptation through danger: she wants to pull the viewer deeper into the ritual while the chamber itself comes alive around her. Wide cinematic anime composition, rich hand-drawn 2D shading, deep crimson and black palette with molten gold sparks, sharp silhouettes, fast readable action, no explicit nudity, no pornographic anatomy, no gore. Pure video, no subtitles, no background music. Stage 1, opening hook: in a ruined ritual hall with towering pillars, floating ash and blood-red sigils burning across the stone floor, the demon girl snaps into frame from darkness with a fast turn, long dark hair whipping behind her, curved horns catching red rim light, slit-pupil eyes locking directly to camera. Environmental pressure: hot air shimmer, circling embers, candle flames bending inward toward the ritual circle. Physical micro-actions: her shoulders roll back, claws open and close once, tail lashes across the floor. Sound design: low ritual pulse, flame rush, sharp heel strike on stone. Camera starts on a fast lateral move then pushes in as she advances, ending on a medium close shot with one glowing hand raised. Stage 2: continue in the same chamber, same identity, same outfit, same direction of movement. She steps out of the circle and spins, dragging a red arc of magic through the air with her fingertips, then lands in a low stance as translucent ribbons and strips of fabric snap around her body in the heat. The sensuality comes from silhouette, motion and eye contact, never exposure. Environmental pressure: sparks burst upward from cracked seams, red smoke curls past the lens, shadows jump across the broken walls. Physical micro-actions: a quick half-smile with a visible fang, chin tilt, chest lift on a sharp inhale, one heel scraping stone as she pivots. Camera orbits around her in a fast half-circle, then whip-cuts with her spin into a tighter angle. End state: partial wings flare open behind her and fill the edges of the wide frame. Stage 3: the ritual accelerates into a kinetic anime action beat. She runs two steps across the glowing floor, leaps, twists mid-air and drives both hands outward, causing talismans and crimson seals to explode into rotating rings around her. Her hair, tail, sleeves and ribbons trail with strong directional motion. Environmental pressure: the chamber wind surges outward, ash spirals upward in columns, molten red cracks race across the floor. Physical micro-actions: fingers flex like she is pulling invisible threads, eyes widen with hungry delight, back arches as the magic peaks. Sound motif: layered pulse, fluttering talismans, one hard bell strike as she lands. Camera pulls back for full-body readability during the jump, then dives toward her landing through the ring of glowing seals. End state: she lands on one knee, then rises slowly with ritual marks glowing at her throat and wrists. Stage 4, climax and release: she strides straight toward camera with dangerous confidence while the entire chamber rotates slightly around her in parallax, then stops just short of the lens and leans in. She lifts two glowing fingers under her lower lip, gives an intimate, threatening smile, and red sigils burst open behind her like demonic wings. Exact on-screen text appears only at the final beat in elegant crimson calligraphy at lower center: "Enter if you dare". Environmental pressure: embers surge past in one final blast, then the air drops into sudden stillness. Physical micro-actions: eyelids lower, lips part slightly, one claw relaxes against her collarbone, tail coils behind her ankle. Camera becomes an aggressive final push into an extreme close portrait, then holds for the last image. Final image: her glowing eyes and dangerous smile dominate the frame while the crimson sigils burn behind her in perfect symmetry. Keep the motion readable, the staging wide-screen, the camera motivated by her turns, leaps and advances, and the tone hypnotic, dangerous, anime-beautiful and erotic-adjacent without explicit sexual content.
Seedance 2.5 — Use the uploaded reference image as the exact character reference. Preserve her facial identity,
Seedance 2.5 · 16:9 · 15s
Use the uploaded reference image as the exact character reference. Preserve her facial identity, grey-green eyes, skin tone, hairstyle, makeup, body proportions, and overall appearance throughout every shot. She has long dark hair in a high ponytail and wears a fitted black long-sleeve athletic crop top, high-waisted black leggings, clean white sneakers, and a white towel around her neck. Maintain perfect character consistency.

Create a realistic late-night fitness vlog filmed on a handheld smartphone with a subtle vintage DV/camcorder aesthetic. The gym is quiet and nearly empty, with large mirrors, dumbbell racks, benches, stretching mats, warm overhead lighting, and realistic reflections.

The video begins with an arm's-length selfie shot as she walks into the gym, slightly breathless but smiling at the camera. She casually pans across the empty gym before returning to her face.

She stops near a mirror, fixes her ponytail, gives herself a playful look, then picks up dumbbells. Cut to a fixed external angle as she performs controlled shoulder presses with realistic form and natural breathing.

Return to handheld selfie footage as she walks with the towel around her neck, looking tired but satisfied. She grabs a water bottle, takes a drink, and smiles at the camera.

Cut to a wider fixed shot where she performs a quick stretch and playful mirror pose. Finish with a close selfie as she brushes back a loose strand of hair, smiles, waves toward the camera, and walks toward the gym exit.

Camera: realistic handheld shake, slight framing imperfections, delayed autofocus, occasional motion blur, subtle exposure shifts, natural jump cuts, no visible camera or camera setup.

Style: authentic creator vlog, glamorous fitness influencer aesthetic, realistic body movement, natural expressions, soft DV texture, subtle tape noise, warm cinematic lighting, shallow depth of field, photorealistic, 4K HDR, 16:9.

Audio: natural gym ambience, footsteps, dumbbells, breathing, water bottle sounds, soft room tone. No background music, no subtitles, no captions, no logos, no watermarks.

Negative Prompt: No face changes, identity drift, outfit changes, extra people, duplicate limbs, distorted hands, broken gym equipment, impossible movements, artificial skin, excessive beauty filters, flickering, cartoon style, visible camera.
Seedance 2.5 — Generate a 30-second, real-time, uncut adventure action film shot in one continuous take. The ma
Seedance 2.5 · 16:9 · 15s
Generate a 30-second, real-time, uncut adventure action film shot in one continuous take. The main subject is reference character @1, wearing reference costume @2, navigating four trap zones in an ancient underground temple to retrieve a modern bottled perfume. [Global Settings] @1 provides character reference: facial features, long curly brown hair, red lipstick, and accessories. @2 provides costume reference: white lace bodysuit, high collar, white belt with bow, translucent cape, and white high heels. Environment is a mossy, crumbling ancient ruin. Lighting uses orange torchlight and cold skylight beams. Visual style is cinematic adventure film texture with shallow depth of field and natural motion blur. Camera language is a 30-second Steadicam long take, moving from 3/4 rear follow to low-angle profile and finally to front mid-close up. [Timeline Breakdown] 0-5s: Darts trap. She steps on a loose plate, dodges darts by rolling. Her white gloves get dusty. 5-12s: Underground waterway stepping stones. She jumps across six sinking stones. The cape hem gets wet. 12-18s: Pendulum blade corridor. She times her sprint past swinging stone blades, leaning back to narrowly avoid the last one. 18-24s: Collapsing bridge sprint. The stone bridge falls behind her as she sprints and leaps to the other side. 24-30s: Perfume altar. She reaches the altar where a modern glass perfume bottle sits illuminated by a beam of light. Torchlight reflects off the glass. No studio lighting.
Seedance 2.5 — Cinematic 30-second video of a glamorous blonde woman driving a classic dark-green 1960s Ford Mu
Seedance 2.5 · 16:9 · 30s
Cinematic 30-second video of a glamorous blonde woman driving a classic dark-green 1960s Ford Mustang coupe through a tree-lined city avenue at golden hour. She wears a patterned silk headscarf, large brown sunglasses, a brown blazer, and large statement rings (one yellow diamond, one emerald).

Stylish slow-motion tracking shots of the car speeding past buildings and traffic. Close-up of her focused face, sunglasses catching the golden light. Close-up of her hands on the wooden steering wheel, rings glinting. Rear-view mirror reflection of her eyes. Her jeweled hand resting on the car door, wind catching a loose strand of hair from the headscarf.

A red traffic light glowing as the Mustang idles, engine humming, heat shimmer off the hood. A motorcycle with a passenger weaves through traffic beside her. Wide low-angle shot of the Mustang's grille and headlights as it accelerates past camera. A bearded man in a black beanie stands on the sidewalk, watching the Mustang drive away into the sunlit street, his gaze lingering.

New scenes: A brief shot of pedestrians and café awnings blurring past in the background bokeh. An aerial/drone-style shot pulling back to reveal the full tree-lined avenue as the Mustang weaves gently between light and shadow from the trees. A final slow-motion hero shot — the Mustang turning a corner into direct sunset flare, silhouette and lens flare dominating the frame, her profile briefly lit gold before the light overtakes the frame.

Warm sunset lighting, lens flares, motion blur, elegant and mysterious atmosphere, high-end commercial style, anamorphic lens feel, shallow depth of field, watermark-free.
Seedance 2.5 — 高品質な日式アニメ(Anime)調の動画で、「作画崩壊(Sakuga)」レベルの流れるような戦闘シーンが描かれます。

00:00–00:00(冒頭フレーム:印を結ぶ手元アップ)

0
Seedance 2.5 · 16:9 · 15s
高品質な日式アニメ(Anime)調の動画で、「作画崩壊(Sakuga)」レベルの流れるような戦闘シーンが描かれます。

00:00–00:00(冒頭フレーム:印を結ぶ手元アップ)

00:00–00:03(法陣生成・地脈覚醒)
カメラ:掌の金の光のアップから急速に引き → 俯瞰ショット
アクション:両手で印を結び、そっと下へ押し当てる。指先が法陣の文様に合わせて微かに震え、大地を撫でるような姿勢で、霊力が掌から地面に染み込んでいく
エフェクト:少女を中心に、金色の平安時代の呪紋の法陣が素早く展開。符文が星河のように流れ、地面のひび割れから金の光が滲み出る。周囲の瓦礫が浮遊回転する。
雰囲気:神聖で威圧的、地脈が応答する

00:03–00:06(式神顕性・オオサンショウウオの霊体降臨)
カメラ:法陣中心のアップ → クイック周回フォロー
アクション:少女は胸の前で両手を素早く組み、印を結びかえる(陰陽師「大鯢召喚印」:両手を重ね、指を絡め、親指を外側に回して環状にする)。その動きは舞のように滑らかで、体は印のリズムに合わせてほんの少し前に傾く
エフェクト:法陣中心から墨緑+金紫の二色の霊力の霞が立ち上る。巨大な日本オオサンショウウオ(大山椒魚)の式神の影がゆっくりと現れる:丸みを帯びた頭部、皺だらけの皮膚、長い尾が揺れ動く。霊力の霞の中に金色の鱗の紋様と水紋の呪印が浮かび上がり、少女の印を結ぶ動作と呼応する。
日本語呪文(清冷で威厳のある声)
『山椒の霊、我に従え!力を授けよ!』
雰囲気:霊体の具象化、契約の締結

00:06–00:09(憑依融合・変身開始)
カメラ:少女の上半身アップ → 顔へ急速接近
アクション:印を結んだ両手を勢いよく胸に押し当てる。オオサンショウウオの影が頭頂と背中から体内に溶け込む。少女の体は少し後ろに反り、霊力が流れ込むのを感じながら、印を結んだ姿勢を融合完了まで保つ。
エフェクト:金紫+墨緑の流れる光が四肢を這う。制服の上に巫女の紋様が浮かび上がり、スカートの裾は軽やかな動きやすい戦闘スカートへと変わる。髪は引き締まったツインテールに束ねられ、全身に水紋状の霊力の障壁が形成される。
ディテール:顔に霊力の輝きが走り、目の表情は沈静から鋭さへ変わる。目じりにオオサンショウウオの霊紋が浮かぶ。
雰囲気:魂と魂の融合、力の覚醒

00:09–00:12(最終変身・短刀顕現・敏捷体完成)
カメラ:見上げる全身ショット → 横からのスローモーション静止画
アクション:印を結んでいた両手が突然に分かれる。短剣(ダガー)が霊力の中から形を成し、抜き放たれる。少女は体を横向きにし、旋回するように片脚で地面を踏み、もう一方の脚を後ろへ軽く上げる。手首を返し、刃を握る構えを取り、姿勢は猫のように軽やか。
エフェクト:短剣は金紫+墨緑の流れる光を放ち、刀身にはオオサンショウウオの鱗紋と巫女の符文が刻まれている。衣の端は風に激しくなびき、背後にはオオサンショウウオの影が敏捷な戦闘霊紋となって付着する(背中の鱗の紋様、手首の水紋など)。
スタイリング:ボディラインにフィットした巫女戦闘服。腰帯に短剣を挿している。動作は軽やかで鋭く、巫女の神聖さと暗殺者の敏捷さを兼ね備える。
雰囲気:変身完了、全戦闘力解放

00:12–00:15(準備完了・初戦予行)
カメラ:360°周回アップ → 短剣を正面で固定撮影(GoPro風)
アクション:少女は手首を軽くしならせ、短剣で鋭い水紋の光弧を描く。爪先で地面を軽く蹴り、狩猟の準備体勢を取る(重心を低くし、膝を少し曲げ、オオサンショウウオの待ち伏せ姿勢のように)。
エフェクト:短剣の光弧が金緑色の水紋の残像を残す。全身の霊力は安定し始め、法陣の微かな光は消えていく。
雰囲気:時を待ち、爆発寸前の静けさ。

要求:60フレーム、四周年OADにふさわしい完璧な品質

Other ByteDance Seed models

Seedance 251 creditsSeedream 4PhotoSeedream 5 ProPhoto

All models →