"Generate a 28-second hyper-realistic video, in landscape composition and 4K resolution, with a seaside couple’s MV‑level authentic texture. {Camera shooting pixel parameters: 1.6}, natural film grain: 2.5, {realism parameter: 1.6}. The lighting transitions are smooth and natural, (each character’s facial features are 100% based on the uploaded reference images—Image 1 and Image 2 represent Character 1 and Character 2, respectively), and the human body proportions are perfectly accurate. Overall, this is a romantic short film set during the golden hour by the sea, narrated through seamless hard cuts and slow zooms, with no stuttering or frame drops throughout. **In all wide‑angle shots featuring Character 2, the face remains stable and clear, without flickering, distortion, or shifting facial features**, creating a gentle, soothing atmosphere brimming with the relaxed romance of a seaside date.


[Video: 28 seconds]

The setting is a sunny evening at the golden hour along the coast: a weathered gray‑brown wooden pier stretches out toward the sea, its wooden pilings adorned with black tire‑shaped bumpers; in the distance, a red‑and‑white fishing boat rests anchored, while a white lighthouse stands at the far end of the shoreline. The crystal‑clear blue‑green water shimmers with tiny ripples, and the beach below is a warm golden sand dotted with natural footprints. As sunset approaches, soft golden light gently illuminates the characters, lending the scene an ethereal, naturally blurred transparency characteristic of the seaside.

At the opening, Character 1 sits alone on the pier, waiting. With Character 2’s entrance, the narrative gradually unfolds.

All movements feel natural and fluid:

- Character 1: Begins seated on the edge of the pier, hands resting casually on their knees as they gaze out to sea. Upon hearing footsteps, they turn to see Character 2 approaching. In the sandy beach scene, they pivot slightly to focus on Character 2 singing. On the pier, they kneel on one knee, holding a camera to capture Character 2 in the distance. During the sunset beach sequence, they sit side by side with Character 2, gazing at the ocean, their expression tender and attentive throughout.
  
- Character 2: Slowly walks from the far end of the pier—a central figure driving the story forward. Standing on the pier, they face the camera and sing naturally, accompanied by relaxed gestures like adjusting their hair. On the beach, they perform a side‑facing vocal performance, pose gracefully for photos, and during the sunset segment, sit shoulder‑to‑shoulder with Character 1, their demeanor calm and gentle. Throughout wide‑angle and panoramic shots, their face remains crisp and steady.


0–4 seconds: A fixed medium shot looking straight ahead, fully showcasing the wooden pier and the surrounding seascape. Character 1 sits on the pier’s edge, gazing out to sea, then turns to look toward the far end of the pier. Meanwhile, Character 2 slowly walks toward the camera, maintaining a fixed position without movement, establishing the initial meeting between the two.

4–6 seconds: A hard cut to a close‑up of Character 2 from the chest up, against the backdrop of the clear blue‑green sea. Character 2 faces the camera, sings naturally, with hands clasped behind their back, their face sharp, steady, and free of flicker.

6–9 seconds: Another hard cut to a warm golden‑sand close‑up of the pair, sitting side by side on the fine sand. Character 2 sings while turning slightly to lift their chin, one hand lightly resting on their knee as they tap rhythmically. Character 1 turns their head to watch Character 2 intently, bathed in the warm golden glow of the setting sun.

9–18 seconds: Back to the pier, a medium shot of Character 2, standing on the pier and continuing to sing. Their arms rise naturally with the rhythm, making pointing and light tapping gestures, occasionally brushing their hair with a graceful motion. The camera slowly pulls back evenly, transitioning from a close‑up to a full view of the pier. Throughout this gradual pull‑back, Character 2’s face remains clear, stable, and flicker‑free, culminating in a wide‑angle shot showing Character 1 kneeling on the near end of the pier, holding a camera to photograph Character 2, who stands on the far end with one hand on her hip.

18–23 seconds: A hard cut to a medium shot of Character 1 kneeling on one knee, crouching on the wooden planks of the pier, raising their camera to capture Character 2 standing at the far end, posing naturally with one hand on her hip. The deep blue sea and pale blue sky serve as the backdrop, illuminated by soft, warm light.

23–24 seconds: A hard cut to a close‑up of Character 2’s face, arms crossed over their shoulders, their expression serene and gentle, against the dark blue sea. The light grows slightly dimmer as the sun sets, creating a quiet, tranquil mood.

24–28 seconds: A hard cut to a panoramic shot of the sunset beach, where warm orange‑pink hues blanket the sky, waves gently lap at the shore, and the couple sits side by side, gazing out to sea. The scene then shifts to a close‑up of the pair, with Character 2 lifting their chin to look into the distance, while Character 1 gently turns their head to watch them. The sequence concludes with the two sitting together, immersed in a relaxed, romantic moment, gazing peacefully at the ocean.


[Key Highlights]

Throughout the entire 28‑second video, transitions are smooth and natural, with no frame drops or sudden brightness spikes. The scenes shift organically with the narrative: the wooden pier, the blue‑green sea, the red‑and‑white fishing boat alongside the white lighthouse, the warm golden sand beach, and the pink‑orange sunset sky—all rendered under the soft, golden glow of the golden hour. In every wide‑angle and panoramic shot featuring Character 2, their face remains clear and stable, free of flicker, distortion, or shifting facial features.

Only natural actions appropriate for a seaside date are permitted: gentle walking, spontaneous singing, rhythmic gestures, hair‑adjusting motions, taking photos with a camera, and sitting side by side to enjoy the view. No alterations to facial features or body proportions are allowed; significant shifts in character positioning, excessive camera shake, illogical scene transitions, jerky movements, limb deformities, finger distortions, additional unrelated characters, or any display of player controls, progress bars, or screenshot interfaces are strictly prohibited. All elements must preserve the tender, romantic essence of a golden‑hour seaside couple’s short film."

Raindance AI Video Generator

"Generate a 10-second hyper-realistic video with 4K HD cinematic texture, featuring {camera shooting pixel parameters: 1.6}, natural image noise level of 3.0, and {realism parameter: 1.6}. The lighting transitions are smooth, and the human body proportions are perfectly accurate. The protagonist’s facial features, gender, temperament, hair color, hairstyle, makeup, and overall outfit are 100% based on the facial characteristics from @image.

[Video: 10 seconds]

This is a short nighttime atmosphere piece shot entirely in motion, with the camera maintaining a constant speed and parallel movement alongside the moving vehicle—no hard cuts or transitions. The rhythm gradually shifts from an unrestrained celebration at the beginning to a relaxed, open-ended conclusion:

0–2 seconds: The scene opens with a side‑view medium close-up—a matte burgundy boxy SUV cruising through the night, set against a pitch‑black sky dotted with faint city lights. A man in a black shirt sits behind the wheel, gripping the steering wheel firmly as he drives. In the passenger seat, a woman leans half her body out of the window, wearing a black strapless top; her hair is whipped violently backward by the strong wind generated by the car’s motion. With both hands clutching a green champagne bottle high above her head, she strains upward, twisting the cork. As the cork pops, a thin stream of liquid shoots straight into the night sky.

2–5 seconds: The camera continues to follow from the side, keeping pace. She lowers her raised arm, bends down, and twists the bottle vigorously—now the champagne erupts fully, sending a torrent of white foam mixed with liquor soaring upward into a translucent mist. The liquid splashes onto the red car’s body, as well as her chest, arms, and cheeks. Her face radiates excitement and joy, completely unfazed by the droplets clinging to her.

5–7 seconds: Bringing the foaming bottle to her lips, she tilts her head back and drinks deeply, letting the liquid cascade down her jawline and neck. After two sips, she raises her arm and hurls the still‑dripping champagne bottle toward the night sky behind her.

7–8.5 seconds: Immediately after tossing the bottle, she lifts a signal flare gun, extends her arm straight toward the heavens, and pulls the trigger, shattering the champagne bottle mid‑air into tiny shards of glass. Turning back toward the camera, she smiles, raises her hand, and sends a flying kiss, her hair dancing wildly in the breeze while streaks of wine glimmer softly under the dim light.

8.5–10 seconds: The camera begins to pull back smoothly, transitioning from a close-up to a wide shot. The perspective naturally shifts to the vehicle’s rear side, revealing the entire car in motion: the slick nighttime road reflects the car’s headlights, its red taillights glowing brightly, and the wheels kick up a fine mist of water as they roll along. Streetlights line the roadside, stretching into a warm ribbon of light that fades into the distance. As the scene gradually turns to black-and-white, elegant white serif lettering reading “Happy Birthday 28” appears alongside a hollow little heart on the right, blending seamlessly into the frame and harmonizing with the road and the receding vehicle. The car travels steadily along the straight highway, disappearing into the deepening night. The camera lingers, watching until the vehicle vanishes at the end of the lit street, while the text remains crisp and clear, gently concluding the emotional arc."

Birthday Blast

Generate a 15-second hyper-realistic video with 4K, cinematic-quality texture and detail, {camera shooting pixel parameters: 1.6}, natural image noise level of 3.0, {realism parameter: 1.6}, seamless light-and-shadow transitions, and perfectly accurate human body proportions. The protagonist’s facial features, gender, demeanor, hair color, hairstyle, makeup, and overall outfit are 100% based on the facial characteristics depicted in @image.
[15-Second Video]
Camera style: Throughout the entire clip, the camera remains fixed at a level-eye view inside the vehicle, with no scene transitions. There is a subtle, natural shake due to the car’s movement. The frame fully encompasses the driver’s seat, the center console armrest, and the passenger seat area, while the road outside is faintly visible through both side windows. Any scenes featuring the handgun are deliberately blurred and not shown clearly.
Action sequence:
0–4 seconds: Seated in the driver’s seat with the seatbelt fastened, the protagonist uses their left hand to steady the steering wheel and picks up a handgun from the center console armrest. They first gently place the gun on the pink cushion of the passenger seat for comparison, then withdraw it, occasionally glancing out the window with a relaxed, casual expression.
5–9 seconds: Raising their right arm, the protagonist aims the handgun toward the left-side window and pulls the trigger. As the window passes by, a black‑clad figure appears, and the glass shatters into a spiderweb‑like crack. Glass fragments scatter across the passenger side. Instinctively frowning and turning their head to dodge the flying debris, the protagonist maintains the raised‑gun posture, briefly pausing with the muzzle pointed toward the camera before firing several rapid shots, all the while keeping their other hand firmly gripping the steering wheel.
10–15 seconds: The protagonist returns the handgun to its position in front of the steering wheel, their expression shifting from tense back to relaxed. They then flip the weapon over with both hands for inspection, finally resting the gun alongside the steering wheel. Turning their head to look out the window, they regain composure; meanwhile, the cup on the passenger seat and armrest has already accumulated a layer of shattered glass.

Badass

Generate a 10-second hyperrealistic video, vertical composition, 4K HD with fashion-advertising-grade realistic texture, {camera shooting pixel parameters 1.6}, natural image noise 3.0, {realism parameters 1.6}, natural light and shadow transitions, and completely accurate human body proportions. The protagonist is 100% based on @image, retaining the main facial features, skin tone, hairstyle, makeup, and overall temperament without change.

[Video 10S]

The entire sequence consists of medium-close-up shots filmed in tracking motion at the same speed as the moving vehicle, with stable framing and focus constantly locked on the protagonist leaning out of the car window; the background is a snow-covered forest road at night, flanked by coniferous forests blanketed in snow, presenting natural backward motion perspective as the vehicle moves forward, all imbued with a wintry Christmas atmosphere complete with falling snow.

Action sequence (total 10s)

0-2s: The protagonist, wearing a red-and-white plush rabbit-ear headband and a red Christmas tube top with white fur trim, leans her upper body out from the side window of a red truck adorned with colorful Christmas string lights. Initially smiling sweetly, she holds up a green bottle of champagne for display, then swings her right arm to toss the champagne high into the air.

2-6s: The protagonist swiftly draws a pistol wrapped in the same colorful Christmas string lights, extends her arm straight outward aiming at the champagne bottle continuously rising and tumbling in midair, her expression focused as she locks onto the champagne—this section is shown in slow motion.

6-7s: The protagonist pulls the trigger, hitting the champagne in midair; the bottle shatters instantly, sending copious white champagne liquid and foam exploding outward like snowy mist, while golden fragments of the bottle seal and cork scatter in all directions, briefly obscuring part of the scene with diffused liquor vapor.

7-10s: After the liquor mist clears, the density of snowflakes in the air noticeably increases, large snowflakes drifting down onto the red car body, the protagonist's hair, and shoulders; maintaining her pose of extending her right arm while holding the gun, she then bursts into unrestrained laughter, tilting her head back with her mouth wide open, exhibiting a joyful and exuberant demeanor, her body naturally trembling slightly with each laugh.

Christmas

Generate a 13-second hyperrealistic video with 4K HD cinematic texture, {camera shooting pixel parameters 1.6}, natural image noise 3.0, {realism parameter 1.6}, natural light and shadow transitions, and perfectly accurate human body proportions; the protagonist's facial features, gender, temperament, hair color, hairstyle, and makeup are 100% based on @image. 
【Video 13S】
The camera movements and content actions in this 13-second video are organized chronologically as follows:
0-2 seconds: Use a fixed medium shot to show the indoor scene of an ice castle: two child characters dressed as green-clad, red-trimmed Christmas elves, wearing red-and-white striped stockings and pointed elf shoes, each holding a small hammer and standing slightly to the left and right of a tall rectangular block of ice, gently but firmly striking it; sealed inside the ice is a protagonist wearing a burgundy strapless velvet short dress adorned with three-dimensional large floral decorations, with hands on hips, eyes closed, and standing upright.
2-4 seconds: The camera pushes forward for a close-up of the ice block, focusing on the protagonist trapped within; soon spiderweb-like cracks spread across the ice surface, the entire block shatters instantly, shards falling downward, while the protagonist opens their eyes, mouths agape in joyful laughter, simultaneously spreading arms wide to complete the ice-breaking action.
4-7 seconds: The camera first pulls back slightly to a medium-close shot of the protagonist, capturing them laughing with arms outstretched as ice shards continue to fall in front of them; then the camera pulls back to a full shot, where the two little elves on either side of the ice block raise their arms to shield their faces from flying shards, while the protagonist stands atop the scattered ice fragments, still indoors in the ice castle.
7-8 seconds: A bright overexposed white flash completes the transition to an outdoor snowy scene, with a full shot: in the background are snow-covered rolling peaks and Christmas trees strung with warm lights, in the foreground are neatly wrapped gift boxes, and fine snowflakes drift through the air. The protagonist stands at the very center of the group of elves, arms raised high, jumping lightly with joy, while numerous surrounding Christmas elves simultaneously perform waving and stomping dance moves.
8-11 seconds: The camera pushes forward to a medium-close shot of the protagonist, who wears a radiant smile, swaying their body and dancing their arms to the rhythm, with the elves behind them synchronously following along.
11-13 seconds: Two child elves wearing colorful lights on their hats run from both sides of the frame to stand before the protagonist, reaching out to wrap their arms around the protagonist's waist; the protagonist smiles and lowers their head, extending their arms to embrace the two children, as the camera freezes on a close-up of the trio in an embrace, with the remaining elves in the background raising their hands in cheers and snowflakes continuing to fall.

Winter Reunion

"Use @video1 as the primary motion reference. Preserve the original body movements, timing, rhythm, pacing and camera movement as faithfully as possible.



Replace the performer with @image1 Keep his face, identity, body proportions and appearance consistent throughout,[role].



Photorealistic cinematic music-video look. Preserve the original camera movement and framing. Prioritize the original synchronized dance choreography above all other changes. No face morphing, identity drift, distorted bodies, extra limbs, floating feet or random movements."

Cat Dance Clip

"Generate a 15-second hyperrealistic video with 4K, cinematic-quality texture. Camera shooting parameters: pixel resolution 1.6; natural image noise level 3.0; realism parameter 1.6. Lighting and shadow transitions are smooth and natural, and human body proportions are perfectly accurate. The protagonist’s facial features, gender, temperament, hair color, hairstyle, makeup, and outfit are 100% based on the facial characteristics from the provided image.

[Video: 15 seconds]

Art Style: A stylized 3D cartoon rendering in a game‑like aesthetic. Characters adhere to standard adult proportions, outlined with clear, dark contours. Color saturation is moderately high, while skin has a matte, cartoon‑like texture—free of realistic pores, fine lines, or other real‑life textures. Fabrics and accessories feature simplified, game‑style materials. Lighting employs bright, layered chiaroscuro typical of cel‑shading, without hyperrealistic physical light effects. A solid dark background of deep blue mist provides depth, illuminated by a vertical spotlight overhead, with subtle airborne dust particles casting soft volumetric light. While modeling details like skin and fabric textures remain true to reality, the overall presentation conveys a formal, character‑casting‑ready feel, as if introducing a user‑based game avatar.

Camera Logic: Throughout the sequence, the camera moves steadily at a constant speed horizontally to the right, with no shaking or zooming. The user‑based character enters the spotlight area from the right side, pauses naturally, and performs a brief, signature gesture (in this video, waving, giving a thumbs‑up, or raising a hand—all executed in a consistent style).

Costume Guidelines: For the first character introduced, the costume faithfully replicates, at a 1:1 scale, the style, color, cut, and accessory details of the photo uploaded by the user, with no modifications. Subsequent characters adopt costumes, accessories, and styling elements directly corresponding to their respective game roles. Each video showcases 6–8 distinct character outfits, avoiding an overwhelming array of looks, with each segment strictly limited to 14–15 seconds.

0–3 seconds: The camera begins with a steady pan to the right across an empty, deep‑blue misty backdrop. The protagonist—dressed exactly as depicted in the user‑uploaded photo—enters the spotlight from the left, stops naturally, and gently raises a hand toward the camera.

3–8 seconds: The camera continues its smooth pan to the right, revealing a user‑based character dressed as a pirate—complete with red bandana, eyepatch, open‑collar white shirt, black short waistcoat, wide red belt, khaki capri pants, and ankle boots—fully retaining all original costume details. This character steps into the spotlight, raises a clenched fist in a friendly greeting toward the camera.

8–13 seconds: The camera keeps moving steadily to the right, introducing another user‑based character dressed as a vintage rugby player—wearing a wine‑red No. 21 long‑sleeve jersey, khaki retro shorts, an old leather helmet, and knee‑high socks—again preserving every original detail. This character stands in the spotlight, giving a thumbs‑up, mimicking the exact gesture seen in the reference video for that role.

13–15 seconds: The camera halts its pan, and three characters in different outfits stand side by side beneath the spotlight. The scene briefly freezes, concluding precisely at 15 seconds."

Halloween Agent Male

"Generate an 18-second hyperrealistic video, with 4K HD cinematic texture and realism. Camera pixel settings: 1.6; natural image noise level: 3.0; realism parameter: 1.6. Light and shadow transitions are seamless, and human body proportions are perfectly accurate. The protagonist’s facial features, gender, temperament, hair color, hairstyle, makeup, and overall outfit are 100% based on the facial characteristics from the provided image.

[Video: 18 seconds]

The entire sequence employs handheld tracking shots that convey a sense of breath and movement, with the camera naturally jiggling and shaking as if following a running figure. The lighting is dim and oppressive throughout, punctuated only by the flickering warm yellow wall lamps in the old residential corridor. Every corner and the far reaches of rooms remain shrouded in deep, inky shadows. All you hear are the protagonist’s heavy, trembling gasps and the creaking of aged wooden floors under pressure—creating an atmosphere of relentless tension from the very first moment:

0–3 seconds: The shot begins from behind the protagonist’s shoulder. They’re crouching on the wooden floor, their fingertips just brushing against scattered, eerie sketch papers when they suddenly hear an extremely faint scraping sound coming from behind them. Their motion freezes instantly, and the camera pans to reveal a tall, gaunt, strangely disproportionate figure standing at the end of the corridor: wearing a white shirt with a burgundy tie, with unnaturally long arms hanging limply at their sides. Their pale face bears a fixed, exaggerated grin, while their black pupils stare unblinkingly straight toward the protagonist. Their head tilts sharply to one side, defying all anatomical logic, and they don’t even blink.

3–7 seconds: The camera cuts back to the protagonist’s face. Their pupils constrict as they gasp sharply, then they whirl around and sprint into the dark recesses of another room connected to the hallway—a narrow passageway leading deeper into the house, with no exit in sight. The camera closely follows, racing forward alongside them. Bare feet pound against the floor, producing frantic, hurried footsteps, and after taking two steps, the protagonist glances back for the first time.

7–12 seconds: The protagonist keeps pushing forward, plunging deeper into the heart of the house. On either side, half-open doors reveal impenetrable black silhouettes. As they round a corner and look back again, a chill runs down their spine: the strange figure isn’t chasing them at all—it has silently appeared three meters ahead, leaning against the base of the wall, sporting that same tilted‑head grin, blocking their path entirely. Terrified, the protagonist slips, scraping their shoulder against the wall, then stumbles into an even narrower side corridor, continuing their desperate flight deeper still.

12–16 seconds: The camera tilts slightly with the protagonist’s staggered steps, mimicking their oxygen-deprived dizziness. Panting heavily, blood trickling from a recent bump on their nose, they turn back for the third time—and freeze completely: the eerie figure has now pressed itself against the doorframe beside them, its excessively long, pallid arm nearly brushing against their shoulder.

16–18 seconds: With their back pressed against the cold wall and eyes filled with despair, the protagonist looks up. The strange figure stands less than half a meter away, its rigid arm slowly rising, icy fingertips hovering just above the protagonist’s cheek. A faint smudge of dark blood clings to the corners of its grinning mouth. And just as those fingers are about to touch the protagonist’s skin, the screen abruptly goes black, cutting off the scene."

Chasing

"Generate a 15-second hyperrealistic video, with 4K HD cinematic-quality texture, {camera shooting pixel parameters 1.6}, natural image noise level 3.0, {realism parameter 1.6}, smooth light-and-shadow transitions, and perfectly accurate human body proportions; the protagonist is 100% based on the facial features, gender, temperament, hair color, hairstyle, makeup, and clothing style from the @image.

[15-Second Video]

Art Style: A stylized 3D cartoon rendering in a game-like aesthetic—characters are rendered at normal adult proportions, with clear dark outline strokes, moderately high color saturation, and a matte cartoon skin texture, devoid of realistic pores, fine lines, or other real‑world textures. Fabrics and accessory materials are simplified in a game‑style manner, while lighting employs bright, layered chiaroscuro typical of cel‑shading, without ultra-realistic physical light effects. The background is a solid deep blue misty shade, illuminated by a vertical spotlight overhead, with fine dust particles floating in the air and soft volumetric lighting. Skin and fabric textures are modeled with realistic detail, creating an overall formal presentation that feels like a character casting showcase based on the user’s likeness.

Camera Logic: Throughout the entire sequence, the camera moves steadily horizontally to the right at a uniform pace, with no shaking or zooming. The user‑based character gradually walks from the right side into the spotlight area, stops, and performs a brief, natural, signature gesture (in this video, the character’s gestures—waving, giving a thumbs-up, or raising a hand—are kept consistent).

Clothing Guidelines: For the first character introduced in the video, the outfit is an exact 1:1 recreation of the style, color, cut, and accessory details from the user‑uploaded photo, with no modifications whatsoever. All subsequent characters wear costumes, accessories, and styling details corresponding to their respective genders within the game universe. Each video features 6–8 distinct game‑character outfits, avoiding an overabundance of looks; each segment is strictly limited to 14–15 seconds.

0–3 seconds: The camera begins with a steady pan to the right across an empty deep blue misty backdrop. The protagonist (whose outfit precisely matches the user‑uploaded photo) enters from the left side into the center of the spotlight, stands naturally, gently tilts their head, lifts a hand to brush aside hair near the ear, and displays a relaxed, natural expression.

3–8 seconds: The camera continues moving smoothly to the right, introducing a cat‑costumed game character based on the user’s likeness. Only the costume is changed—a black bodysuit, pink cat‑ear headband, gold waist chain, long black boots, and a cat tail accessory—and the character walks in from the right side into the spotlight, standing still and making a playful “cat paw” gesture toward the camera.

8–13 seconds: The camera keeps panning steadily to the right, revealing a Little Red Riding Hood‑costumed game character (white bubble‑sleeve top, black cinched waist, red‑trimmed white pleated short skirt, red cape, red bow tie, black silk stockings and boots). This character steps into the spotlight, lightly grasps both sides of the skirt, and gives a slight bow toward the camera as a subtle gesture.

13–15 seconds: The camera halts its horizontal movement, and three game characters in different outfits stand side by side under the spotlight. The scene briefly freezes for a final touch, bringing the video to exactly 15 seconds."

Halloween Agent Female

"Generate a 15-second hyperrealistic video, in 4K HD with cinematic-quality texture, {camera shooting pixel parameters 1.6}, natural image noise level 3.0, {realism parameter 1.6}, with smooth light-and-shadow transitions and perfectly accurate human body proportions; the protagonist’s facial features, gender, temperament, hair color, hairstyle, makeup, and outfit are 100% based on the characteristics of the @image.

[15-Second Video]

Shooting perspective: Throughout, we use a passerby‑style handheld vlog shot from an eye‑level viewpoint, with subtle natural hand‑shaking to simulate the authentic, on‑the‑spot feeling of tourists casually recording moments—no deliberate camera movements, and the focus always follows the interaction between the protagonist and the ghostly figure.

Scene setting: An outdoor Halloween-themed market in autumn, with cement walkways scattered with withered yellow leaves, white event tents standing along the sides, piles of hay decorating the tent edges, and purple string lights winding around them. Scattered visitors dressed in casual autumn attire—sweatshirts, jeans, sneakers—either stop to watch or stroll by, creating a lively yet eerily Halloween‑inspired daytime outdoor atmosphere, illuminated by soft, natural autumn light.

Character design: The protagonist keeps the original outfit from the uploaded image throughout; the ghostly figure is styled after Michael Myers from the reference photo: tall and slender, wearing a stark white mask with a blank expression, a long black work jumpsuit, red gloves, and holding a prop knife with fake blood trailing from its blade. Movements start out stiff and slow, then become explosively powerful later on.

Shot-by-shot action (full 15-second rhythm):

0–4 seconds: The protagonist stands by the walkway, head down looking at their phone, completely unaware of what’s happening behind them. Meanwhile, the ghostly figure silently and extremely slowly creeps up from the side and rear, leaning slightly forward, their masked face fixed intently on the protagonist’s back. The distance gradually shrinks from 3 meters to half a meter, while nearby spectators hold back laughter as they watch the two.

4–9 seconds: Suddenly realizing something is amiss, the protagonist whips around, eyes widening in shock upon seeing the ghostly figure just inches away. They frantically pull out their phone, raise it toward themselves, point the camera at both themselves and the ghostly figure behind them, awkwardly make a “scissors” gesture in preparation for a selfie, and nervously sidestep slightly to the side. The ghostly figure remains motionless behind them, their mask squarely facing the camera with no reaction.

9–12 seconds: Just as the protagonist presses the shutter button and the phone’s flash flickers on, the ghostly figure suddenly surges forward with incredible speed, raising their prop knife in a mock slashing motion, maximizing the scare factor.

12–15 seconds: Startled into a loud scream, the protagonist nearly drops their phone, spins around, and scrambles forward along the walkway, stumbling and nearly slipping on the fallen leaves. The ghostly figure strides confidently behind them, chasing persistently with exaggerated, mischievous comedic flair. Nearby spectators burst into laughter, while the camera shakes slightly following the duo’s frantic movement, finally settling on a close-up of the protagonist fleeing in panic, with the ghostly figure hot on their heels, bringing the scene to a close.

Overall vibe: A genuine street‑prank feel—building eerie tension in the opening moments, delivering maximum fright during the sudden rush, then shifting to lighthearted comedy in the chase. No bloody or terrifying content—just a fun, interactive record of a Halloween event."

Say Cheese

"Generate a 15-second hyperrealistic video, with 4K HD cinematic texture, {camera shooting pixel parameters 1.6}, natural image noise level 3.0, {realism parameter 1.6}, smooth light-and-shadow transitions, and perfectly accurate human body proportions; the protagonist’s facial features, gender, temperament, hair color, hairstyle, makeup, and clothing style are all 100% based on the characteristics of the @image.

[15-Second Video]

A 15-second horror prank scene set in an amusement park, shot from a fixed medium-long camera position (simulating a handheld perspective or roadside recording without significant camera movement, enhancing the immersive realism of the prank). The overall tone is overcast and cool gray, accompanied throughout by subtle fog effects:

0–4 seconds (introduction): The camera pans across the main asphalt pathway of the amusement park, revealing the interwoven steel framework of a wooden roller coaster in the background, with green trees, red flowerbeds, and benches clearly visible on either side. Scattered visitors sit or stroll along the path. From the far end of the walkway, the protagonist slowly emerges from the haze: pale ghostly makeup—deep black smudges covering large areas around the eye sockets, exaggerated upward‑slanting cracks at the corners of the mouth—disheveled hair, and a wrinkled black outfit stained with fake dirt. Protective gear hangs askew, a long prop knife dangles loosely from the back, steps are stiff and jerky, knee lifts appear unnatural, the head tilts slightly to one side, eyes stare straight ahead, hands hang limply at the sides, fingers occasionally twitch unnaturally. At this point, nearby visitors remain unaware.

4–7 seconds (triggering the scare): About five meters from the camera, the protagonist suddenly freezes, then mechanically turns their neck toward two passersby sitting on a bench to the right. With a forceful, awkward twist of the waist, arms slowly extend outward to their maximum range, before abruptly bending forward nearly 45 degrees, head tilted, staring intently at the pedestrians. The motion feels like a puppet whose joints have seized up.

7–11 seconds (panic erupts): Upon making eye contact, the pedestrians’ faces turn deathly pale, they let out shrill screams, leap off the benches, and bolt away in panic. Nearby visitors are also swept up in the chaos, scattering to either side; some even slip and nearly fall. The protagonist suddenly straightens up, charging toward the fleeing pedestrian with uncoordinated, bizarre strides. Arms swing wildly as they run, feet stumble, nearly tripping themselves, while reaching out with exaggerated, contorted gestures—though joint movements still conform to human anatomy.

11–15 seconds (comical chase): After taking three steps chasing the escaping pedestrian, the protagonist abruptly stops, spins sharply, and lunges toward another passerby trying to slip around them. The second person lets out a terrified scream and darts into a dim corner beneath the roller coaster, pursued closely by the protagonist. During the chase, the character deliberately flails arms and legs in a comically clumsy manner. Some bystanders laugh, others pull out their phones to film. Finally, the protagonist charges after the pedestrian into the shadowy space under the steel framework, where wisps of smoke drift through, partially obscuring the view and leaving behind a chaotic yet amusing afterglow of the pursuit.

Run From Me

"Generate a 10-second hyperrealistic video, with 4K HD cinematic texture, {camera shooting pixel parameters 1.6}, natural image noise level 3.0, {realism parameter 1.6}, smooth light-and-shadow transitions, and perfectly accurate human body proportions; the protagonist’s facial features, gender, temperament, hair color, hairstyle, makeup, and clothing style are 100% based on the facial characteristics in @the image.

[Video: 10 seconds]

The entire sequence is shot from a first-person handheld selfie perspective, capturing the authentic feel of a mobile phone night-time documentary-style snapshot, with natural digital grain in the shadows, slight motion blur on moving figures and smoke, and no subtitles throughout:

0–3 seconds: The camera faces the protagonist directly, clearly capturing their facial features. Holding the phone straight toward the lens, they initially stare expressionlessly. In the background is a Halloween parade at a nighttime American small-town commercial street intersection: in the foreground, a white zebra crossing on the asphalt road creates a sense of depth; in the middle of the street, layered stage smoke rises, with the smoke near the ground bathed in warm red by the float’s red spotlight, while the rising smoke takes on a cool bluish-green hue under the overhead teal spotlights. On the left, a black pickup truck float can be seen, alongside a small wooden cabin illuminated by red lights; silver temporary guardrails line both sides of the street, packed with people holding phones to watch, with multi-story red-brick buildings featuring pointed roofs and warm yellow lights spilling from their windows. Purple and orange Halloween decorative lights dot the streets, creating a lively yet magically eerie Halloween atmosphere. A few meters behind the protagonist stand two supporting characters dressed as ghosts, styled in spooky fashion: one is a skeletal zombie with disheveled gray-white hair and a tattered yellow burlap smock, revealing a grimacing skull face with sharp teeth and dangling scraps of cloth; the other is a skeletal ghoul wearing a ragged dark-brown robe, exposing a pale skull head and bone-like arms. At first, both stand rigidly, their gazes fixed on the camera.

3–6 seconds: The protagonist suddenly frowns at the camera and lets out a loud shout aimed at the two supporting characters, attempting to intimidate them. Hearing the shout, the two ghostly figures behind the protagonist exchange a quick glance, their movements stiff and momentarily stunned, then immediately turn back to fix their gaze intently on the protagonist, leaning forward slightly with their gaunt arms raised in a posture as if about to chase after them, while colorful smoke partially obscures their tattered garments.

6–10 seconds: Catching movement out of the corner of their eye, the protagonist abruptly turns their head, widening their eyes in an exaggerated expression of surprise and panic, then instantly spins around and starts sprinting forward, phone held high. The camera shakes violently with each step, producing natural motion blur typical of running, while the night-shot grain becomes even more pronounced with the movement. Layered red-and-blue stage smoke drifts past, and the two ragged ghostly figures stride rapidly behind, their skeletal arms reaching forward. Nearby parade performers and spectators behind the guardrails all turn to watch. As the protagonist runs, they keep glancing back, their face showing a mix of panicked confusion and playful mischief—creating a consistently humorous, mischievous holiday prank vibe."

Trick

"Generate an 8-second hyperrealistic video with 4K HD cinematic texture, {camera shooting pixel parameters 1.6}, natural image noise level 3.0, {realism parameter 1.6}, smooth light-and-shadow transitions, and perfectly accurate human body proportions. The protagonist’s facial features, gender, temperament, hair color, hairstyle, makeup, and clothing style are 100% based on the characteristics of the @image.

[8-Second Video]

This is an 8-second 4D cinema horror special-effects experience video that fully preserves the authentic theater environment—dark red walls, red-and-black seating—and features a stylized evil spirit with red eyes and long hair, clad in a dirty, old white dress, matched to the 4D seat effects for a complete chain reaction of scares.

From 0 to 1.2 seconds, the camera starts with a close-up of the protagonist holding a white beverage cup with a straw in the front row, quickly pulling back to reveal the full view of the packed auditorium from a fixed over-the-shoulder position in the back row. At this point, the big screen has not yet fully lit up; only scattered small lights above remain on, and the audience is relaxedly waiting, many already pointing their phones at the screen ready to record. A nearby man wearing glasses and a black T-shirt printed with yellow letters leans casually against his seat, creating a calm, laid-back atmosphere.

From 1.2 to 3.5 seconds, the camera stays fixed in the back row, slightly moving forward to focus on the screen. Once the screen lights up, we see a gloomy forest scene shrouded in cold blue mist: twisted, dead trees draped with vines, a damp, muddy path beneath, and deep within the woods stands the evil spirit—its skin pale grayish-blue, eyes glowing with eerie red light, sharp fangs bared, dressed in a tattered, dirty off-white long skirt, long black nails dangling by its side, slowly leaning forward as it moves toward the screen, its messy long hair fluttering with each motion. At this moment, the audience still hasn’t noticed anything unusual, continuing to relax and film with their phones, some curiously peering closer at the screen.

From 3.5 to 6 seconds, the camera remains fixed while gradually zooming in as the spirit approaches, intensifying the visual sense of pressure. The spirit gets closer and closer, its figure steadily growing larger until it nearly reaches the edge of the screen. It raises both hands, its long-fingered palms pressing against the screen as if about to pierce through at any moment. At this point, the 4D seats begin to vibrate at low frequencies, accompanied by faint cool breezes sweeping across the audience. Gradually, viewers start noticing something amiss—their hands gripping their phones tighten slightly, most people instinctively lean back, their relaxed postures suddenly stiffening, their gazes locked firmly on the screen.

From 6 to 7 seconds, the camera rapidly swings forward and downward, creating strong motion blur to simulate the first-person impact of the spirit lunging toward the audience. Visually, the spirit breaks right through the screen boundary: its upper body extends beyond the frame, its pale grayish-blue hand reaching out to rest on the steps of the central aisle, its body crawling along the aisle toward the audience. Its blood-red eyes stand out starkly in the dim theater. Simultaneously, the 4D seats trigger intense vibrations and a sudden forward-leaning effect, paired with piercing, spine-chilling sound effects. The entire audience erupts in startled reactions—everyone jerks backward to dodge, those in the front row instinctively raise their hands to shield their faces, and phone-holding spectators’ wrists tremble uncontrollably. Waves of gasps and screams fill the room.

From 7 to 8 seconds, the camera swiftly cuts back to the opening close-up shot in the front row, returning to the position of the black-clad girl. The scene stabilizes again, but now the spirit has crawled all the way to the front of the aisle, completing the 4D horror effect of “breaking through the screen.” The chain reaction of fright has fully played out: the protagonist in the front row recoils in shock, burying her face behind the arm of the person next to her, while others instinctively raise their arms to shield themselves, realizing it was just a theater special effect. Nearby viewers either cover their mouths in stunned disbelief or can’t help but laugh after being spooked, while those in the back still hold their phones, everyone retaining that tense, startled feeling. Overall, the mood shifts from initial surprise at the sci-fi impact of the scene to genuine terror caused by the spirit breaking through the screen.

The camera quickly cuts back to the opening close-up shot in the front row, settling once more on the black-clad girl. Her shoulders rise slightly as she recovers from the scare, clutching her drink tightly enough to deform the cup. The man beside her, wearing glasses, regains his composure and gently pats her back. Nearby audience members also gradually snap out of their shock—some pat their chests to catch their breath, others review the chaotic footage they captured moments earlier. Just then, the special-effect sequence of the spirit in the aisle winds down, the screen dims gradually, and the 4D seat vibrations come to a complete stop.

From 8 to 10 seconds, the protagonist slowly lifts her head, staring blankly for a couple of seconds before realizing that everything she just saw was purely a 4D effect. She can’t help but chuckle awkwardly, her cheeks still flushed with the lingering blush of fear. She instinctively covers her mouth with her hand, furrows her brow slightly, her eyes still reflecting lingering fear and disbelief. The atmosphere in the room has completely shifted—from the tense, spine-tingling fright of earlier to a lighthearted smile born of being thoroughly impressed by the special effects. And so, the entire video comes to an end.

4D Thrill

Generate a 20-second hyper-realistic video with 4K, cinematic-quality texture. Camera pixel parameters: 1.6; natural image noise level: 3.0; realism parameter: 1.6. The lighting and shadow transitions are smooth, and the human body proportions are perfectly accurate. The characters fully reference the facial features, gender, temperament, hair color, hairstyle, and outfit of Characters 1 and 2 in the provided images, maintaining their form throughout the performance without distortion, with faces always clearly recognizable.

[Video – 20 seconds]

In the opening 0–2 seconds, the scene sets up the stage. The shot begins as a symmetrical medium close-up from a level perspective, with both performers standing directly beneath the microphone. Lead vocalist Character 1 stands front and center, head bowed to adjust, hands resting naturally at their sides in preparation for singing. Character 2 stands slightly to the side of Character 1, turning toward them, right hand raised to lightly touch the chin and clear the throat, left hand hanging casually by the side, waiting to synchronize with the lead singer’s entrance into the rhythm. During this intro pause, the frame remains steady, clearly emphasizing the lead singer’s central position.

From 2–5 seconds, the moment arrives when the beat drops and the performance kicks off. The camera stays fixed on a symmetrical medium close-up, keeping the microphone centered at the top of the frame and Character 1 firmly positioned at the visual center. As the beat starts, Character 1 opens their mouth and launches into the song, immediately entering an immersive rap mode. Their body subtly bounces up and down in time with the drumbeat, while the right hand rises to mark the first strong beat with a forward-pointing gesture. Simultaneously, Character 2 raises their hand, mirroring Character 1’s movements exactly, creating perfect rhythmic synchronization between the two. This seamless coordination draws the audience’s attention squarely onto the lead singer’s performance.

Between 5–9 seconds, the stable vocal section unfolds. The camera remains locked, keeping both performers in frame. Character 1 stands steadfastly in front of the microphone, fully focused on singing, their head nodding naturally with the flow. Hands alternately execute finger‑pointing and interlaced gestures typical of rap, while the body sways gently—a true focal point of the vocal delivery. Character 2 stands beside the lead singer, using alternating open‑palm, flipping, and sweeping motions to complement the rhythm, their body shifting slightly left and right in sync with the beat. Occasionally, they turn to glance at the lead singer, engaging in subtle interaction without stealing the spotlight—using expressive gestures to enhance the steady output of the main performer. The duo’s chemistry feels effortless and intuitive.

At 9–13 seconds, a pivotal moment of movement occurs. The camera maintains its fixed medium close-up. Character 2 introduces dance‑like body language: hands clasped together in front of the chest, torso pivoting slowly to the right—away from the camera—until their back is turned. Arms swing gracefully in time with the beat, feet shuffle slightly, revealing the elegant lines of their back, while the body undulates softly in response to the rhythm. Meanwhile, Character 1 remains facing the microphone, holding the iconic “gun‑finger” gesture formed by index and thumb, eyes fixed firmly on the mic, completely unfazed by Character 2’s turn. Character 2’s pivot aligns precisely with the eighth‑beat transition, breaking the previous static composition and injecting dynamic variation into the scene, yet still preserving the overall rhythmic stability.

From 13–16 seconds, the flow shifts dramatically during an interactive passage. Character 2 turns back to face Character 1 and the camera, extending both index fingers toward each other, then simultaneously raising them to their lips in a hushed “shh” gesture—perfectly timed to coincide with the heavy beat marking the transition. After lowering their hands, Character 2 performs a light forehead‑wiping motion reminiscent of a salute, while Character 1 spreads their hands wide in a relaxed gesture. Both wear expressions of ease and confidence, exuding a sense of fluidity and mastery over their performance. These synchronized gestures further reinforce the tight bond between the pair, pushing the emotional intensity of the segment to a small peak.

Finally, from 16–20 seconds, the performance winds down. The camera begins a slow optical zoom-out, gradually revealing the full yellow platform and the black speaker in the lower-left corner, returning to the symmetrical framing established at the outset. Character 2 lifts their hand toward the overhead microphone before letting it fall, while Character 1 makes gentle, open‑palm gestures with their hands. The two slightly adjust their positions, facing each other, and respond to the final drumbeats with subtle body bounces. Character 2 gives their hands and clothes a quick sweep, then looks up at Character 1. Together, they settle into place directly beneath the microphone, bringing the performance to a calm and graceful conclusion. The zoom-out action synchronizes perfectly with the closing moments, completing the circular structure of the piece and allowing the rhythm to fade naturally from its climax.

Throughout the video, the cinematography remains deliberately restrained, eschewing unnecessary camera movements or cuts in favor of serving the core rap performance. Every gesture, turn, and interactive move aligns precisely with the beat, and both performers maintain a subtle, continuous knee‑bouncing rhythm—a quintessential studio cypher vibe. With Character 1 delivering steady vocals and Character 2 providing supportive accompaniment, the division of labor is clear, the coordination seamless, and the composed framing coupled with meticulous timing ensuring that this 20‑second piece flows effortlessly yet retains a relaxed, unhurried quality.

Bar Rap

Use @video1 as the primary reference. Strive to faithfully preserve the original shot’s scenes, movements, timing, rhythm, speed, and camera techniques. Keep the characters’ costumes and styling unchanged, and maintain their original makeup, skin tone, and clown‑style makeup.

Replace the central subject in the frame with @image1, ensuring that the subject’s features, facial details, overall appearance, and body shape remain identical throughout. [Maintain the same number of subjects as in @video1; duplicates are permitted, and all subjects must share identical styling.]

Create an MV with a realistic cinematic aesthetic. Retain the original camera movements and compositions. Prioritize preserving the original synchronized choreography above all other modifications. Do not allow any deformation of the subject or random, unnatural movements.

Seedance 2.5 AI Video Templates

Explore AI video templates powered by Seedance 2.5.