Pricing

My Creations

Home

Template

Blog

Assert

Resources

Motion Control

Video Tools

AI Tools

Prank Haircut

Lover Tattoo

Muse Tattoo

Video

AI Anime Filter

AI Manga Filter

AI Watercolor Generator

Image

AI Effects

Generate a vivid video with your inspirational prompt.

Text to Image

Magic image-to-video AI generator brings your still photos to life.

Image to Image

Nano Banana 2.0

GPT Image 2.5

Image Models

Ai Image

AI Image

Generate a vivid video with your inspirational prompt.

Text to Video

Image to video

Edit Video

Set the start and end. AI creates the story in between.

Frame to video

Seedance 2.0

Kling Omni 3.0

Seedance 2.5

Video Models

AI Video

Menu bar

Image to Video

Video Templates

AI Dance Generator

页脚-AI Video

Kling

Kling 3.0 Omni

页脚-Video Models

Image Templates

Birthday Photos

Product Images

ID Photos

页脚-AI Image

GPT Image 2

Nano Banana 2

Nano Banana Pro

页脚-Image Models

页脚-Resource

AI Tattoo Generator

AI Mugshot Generator

AI Character Generator

Ghibli AI Generator

页脚-Use Cases

Use @video1 as the primary motion reference. Preserve the original STORM II synchronized choreography, group formations, body movements, timing, rhythm, pacing and camera movement as faithfully as possible. Do not invent new choreography, walking or posing.

Replace the central performer with @image1 Keep his face, identity, body proportions and appearance consistent throughout. He wears a black Yakuza-inspired suit, black shirt and black sunglasses.

Use @image2 as the location and group-style reference. Rebuild the scene on the Japanese temple staircase at night, with wet stone steps, warm lanterns, mist and cinematic reflections. The surrounding dancers wear matching white suits with black shirts and black ties while maintaining the original formations and synchronized movements.

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.

Genjutsu

"Generate a 15-second hyper-realistic video, with 4K HD cinematic texture and realism. Camera settings: pixel resolution 1.6; natural image noise level 3.0; realism parameter 1.6. Lighting and shadow transitions are seamless, and human body proportions are perfectly accurate. The protagonist’s facial features, gender, temperament, hair color, hairstyle, makeup, and outfit are all 100% based on the reference image.

[Video: 15 seconds]

The entire scene was shot on a sunny day under natural light, set in an old Southern European stone‑paved street. The pacing alternates between romantic, mythic tones in the first half and fast‑cut, exaggerated comedic contrasts in the second:

0–3 seconds: Establishing the tone. A fixed medium‑shot from eye level focuses on a stone terrace, where the male lead—Cupid, adorned with enormous white wings—leans against the railing, wearing a sleeveless white gown. Initially languid with his eyes half‑closed, he lifts them to catch sight of the female lead below, revealing a mischievous smile. A quick cut to a one‑second overhead shot reveals the female lead, dressed in a short, pleated silk dress in pale white, walking along the cobblestone street, her back turned to the camera.

3–7 seconds: Locking onto flight. A rapid zoom-in brings Cupid into close-up; he straightens up, draws back his golden bow, and, with a playful glint in his eye, aims toward the female lead. The camera then follows her in a swift, low‑angle tracking shot as Cupid spreads his white wings, leaps off the terrace, and glides down to land on the street behind her.

7–12 seconds: The twist unfolds. Back to a medium‑shot from eye level, the camera faces the female lead directly. After landing, Cupid stands behind her, bow raised and aimed at her back. Suddenly, she stops in her tracks and whips around. Cupid maintains his confident stance, drawing his bow, when she raises her hand—and with a decisive swing of the bag she’s carrying—smacks him squarely across the face. Stunned, Cupid drops his golden bow with a clatter onto the cobblestones.

12–15 seconds: The payoff. Cupid collapses backward, sprawled out on the pavement with his white wings splayed wide. The camera cuts to a close‑up of the female lead’s face; expressionless, she lowers her head to glance briefly at the fallen Cupid, calmly adjusts her hair, picks up her handbag, turns, and walks away without giving the man on the ground another thought. The scene fades to black in a rapid cut.

Overall, the video juxtaposes the classic romantic expectation of Cupid shooting his love arrow in the first half with the unexpected, slapstick outcome of being knocked down by the female lead’s bag in the second half, creating a sharp comedic contrast. With minimal camera movement, the rhythm of rapid cuts amplifies the humor."

Cupid

"Generate an 8-second hyper-realistic video, with 4K HD cinematic-quality texture and realism. {Camera shooting pixel parameters: 1.6}, natural image noise level: 3.0, {realism parameter: 1.6}. The lighting and shadow transitions are natural, and the human body proportions are perfectly accurate. The protagonist’s facial features, gender, temperament, hair color, hairstyle, makeup, and outfit are all 100% based on the facial characteristics from @image.

[Video: 8 seconds]

Content breakdown for the first 5 seconds (0:00–0:05):

0:00–0:01  
Shot: A fixed, over-the-shoulder close-up shot, with the white-clad man on the right side of the frame as the foreground. The focus is entirely on the woman opposite him, against a backdrop of warm yellow streetlights in a nighttime outdoor setting—softly blurred light spots—accompanied by a very subtle sense of natural hand-held movement and breathing, without any zooming, panning, or tilting.  

Character: The woman lifts her eyes gently toward the man across from her, her eyes already brimming with tears. Her lips part slightly, as if she’s about to speak.  

Atmosphere: The soft glow of warm streetlights at night, combined with the blurred light spots, creates a poignant yet tender farewell ambiance.  

0:01–0:02  
Shot: Maintaining the same over-the-shoulder close-up perspective, the focus remains locked on the woman’s face.  

Character: She first lowers her eyelids, curves her mouth into a bittersweet, forced smile, her eyelashes trembling slightly, while tears glisten in her eyes. Then, she raises her gaze again, her lips parting ever so slightly, shifting from feigned calmness to unmistakable sorrow.  

0:02–0:03  
Shot: The camera pulls back slightly, still maintaining an over-the-shoulder view, revealing more details of the woman’s sari-clad shoulders.  

Character: She furrows her brow gently, closes her eyes, and lets a tear slide down her reddened cheek. Her head tilts back slightly, her shoulders quiver ever so subtly, as she struggles to hold back her sobs—the depth of her grief beginning to surface.  

0:03–0:04  
Shot: A quick cut to a side-facing close-up, focusing on her tear-streaked profile, with the background further darkened and blurred.  

Character: She turns her face away, lowering her head to avoid meeting his gaze. More tears cascade down her cheeks, her lips pressed tightly together, her jawline tense as she fights to suppress her sobbing. Accessories sway gently with each slight nod of her head.  

0:04–0:05  
Shot: Switching to a stable, over-the-shoulder medium shot from behind, capturing her silhouette.  

Character: No longer turning back, she pivots away from the man standing before her, walking slowly forward along the nighttime street. Her long hair falls gracefully behind her, her clothing gleaming softly in the night, her figure looking forlorn as she completes the act of turning away—a final goodbye.  

Atmosphere: The scene is set against the backdrop of streetlights lining the road and the hazy pavement, where cool nighttime tones intermingle with warm golden hues, intensifying the profound sadness of this farewell."

Crying Face

"Generate a 12-second hyper-realistic video, with 4K high-definition cinematic-quality texture. Camera pixel parameters: 1.6; natural image noise level: 3.0; realism parameter: 1.6. The lighting and shadows transition naturally, and the human body proportions are perfectly accurate. The protagonist’s facial features, gender, temperament, hair color, hairstyle, makeup, and outfit are all 100% based on the characteristics depicted in the provided image.

[Video: 12 seconds]

0:00–0:02  
Shot: Handheld medium-close-up, with the camera positioned slightly to the front and diagonal of the subject, featuring subtle natural hand‑shake. The focus remains locked on the protagonist, while the background—dinner guests and table settings—is softly blurred, without any zooming, panning, or tilting movements.  

Character: Initially seated sideways, facing left, appearing to engage with someone nearby. Then, the protagonist slowly turns their head toward the camera, meeting the viewer’s gaze with a gentle, faint smile. Dressed in a sparkling diamond‑encrusted strapless evening gown, with loose, naturally flowing hair and exquisite makeup.  

Ambient atmosphere: A formal dinner setting, bathed in warm candlelight. The table is adorned with white floral arrangements, tall candle holders, and glass wine goblets. Guests are elegantly dressed, creating an overall ambiance that is both refined and relaxed, exuding the lively yet unhurried spirit of a banquet.  

0:02–0:04  
Shot: Maintaining the same medium-close-up perspective, with no significant movement—only a slight, natural handheld shake.  

Character: Raising the right hand, the protagonist performs a playful “kiss‑to‑the‑air” gesture—lifting the palm gently to the lips before sending it outward—while slowly turning their body toward the camera. Transitioning from a side‑seated position to sitting directly facing the lens, their smile broadens, revealing teeth, radiating sweetness and warmth.  

Ambient atmosphere: Warm yellow candlelight flickers gently, with guests chatting and interacting around them. The mood is light, joyful, and deeply inviting, emphasizing the cozy, convivial essence of the gathering.  

0:04–0:08  
Shot: Continuing the medium-close-up view from eye level, with minimal natural follow‑up camera movement as the subject subtly sways. The framing stays focused on the upper body.  

Character: Fully turned to face the camera, maintaining a radiant smile. Their gaze briefly drifts to the side before returning to the lens. Both hands rise to the sides, gently swaying their shoulders in time with the music, with small, fluid arm movements. Fingers relaxed and exposed, showcasing delicate French manicure nails, along with slender bracelets and rings. They move gracefully with the rhythm, appearing completely at ease and savoring the moment.  

Expression: A bright, confident smile, with a relaxed and natural demeanor, fully immersed in the laid-back, festive atmosphere of the banquet.  

Ambient atmosphere: Soft, warm lighting envelops the scene, with guests moving and conversing naturally. Candlelight and sequined gowns complement each other, creating a romantic, elegant ambiance brimming with the joyous spirit of celebration.  

0:08–0:12  
Shot: Keeping the handheld medium-close-up perspective, with stable positioning and only subtle, lifelike camera shake.  

Character: Pausing the subtle hand movements, the protagonist lowers their hands naturally, leaning forward slightly. Then, tilting their head gently to one side, they offer a tender, sweet smile, engaging warmly with the camera. Their posture is graceful and relaxed.  

Expression: Eyes soft and assured, a broad, open smile conveying a sense of comfortable acceptance of age and a positive, carefree attitude toward enjoying the present moment.  

Ambient atmosphere: The warm candlelit ambiance persists, with guests moving naturally around the scene. Overall, the mood remains warm and elegant, radiating strong emotional appeal.  

Overall shot characteristics: Throughout the sequence, there are no dramatic camera moves—primarily fixed handheld close-ups capturing everyday authenticity. Visual transitions are driven by the subject’s own head turns, body rotations, and subtle gestures. Framing consistently focuses on the upper body in medium-close-up, adopting an eye-level perspective that mirrors a natural observer’s viewpoint, faithfully documenting the relaxed, confident state of the individual within the banquet setting.

Glimpse

"Generate a 20-second hyper-realistic video, with 4K HD cinematic texture and realism. Camera settings: pixel resolution 1.6; natural image noise level 3.0; 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 outfit are all 100% based on the reference image.

[20-Second Video]

This is a studio‑shot cypher‑style rap video filmed in one continuous take with a fixed camera position—no cuts or scene changes. Only optical zoom is used to adjust the shot composition, resulting in a clean, crisp visual style. The setting is a standard solid‑color studio: a warm orange matte backdrop fills the entire background, while a raised bright yellow performance platform stands in front. In the lower left corner of the platform sits a small black monitor speaker. A large silver‑diaphragm condenser microphone hangs directly above the center of both performers’ heads, serving as a constant visual anchor. The overall lighting is even, soft, and warmly toned. The duo’s black‑and‑purple ensemble creates a striking contrast against the warm orange and bright yellow backdrop, ensuring that all visual elements focus attention squarely on the performers themselves.

From 0 to 2 seconds, the video establishes its opening phase. The shot begins at eye level, capturing a full‑frame panorama with perfect symmetry. Both performers stand beneath the central microphone on the platform. Performer A wears a black short‑sleeved hooded top paired with white drawstring black wide‑leg pants, accessorized with earrings and a ring. They look down at their wrist, right hand resting atop the left, assuming a preparatory pose—checking the time and adjusting their sleeve. Performer B dons a black‑and‑purple geometric‑patchwork tracksuit, adorned with a silver bracelet and multiple rings. Turning slightly toward A, they raise their right hand to their mouth to clear their throat, while their left hand hangs naturally at their side. Both are waiting for the beat to drop; during this intro pause, the frame remains steady, clearly signaling to viewers that the performance is about to begin.

Between 2 and 5 seconds, the moment arrives when the beat kicks in. The camera slowly zooms in using optical zoom, eventually settling on a medium close‑up shot, keeping the microphone centered at the top of the frame. As the zoom progresses, Performer A lifts their head, closes their eyes, and immediately enters an immersive rapping state, their body subtly bobbing up and down in time with the drumbeat, right hand lightly tapping their waist. Meanwhile, Performer B turns toward the camera, extending their right index finger toward the lens to mark the first heavy beat, nodding relaxedly. At the same instant, Performer A mirrors this gesture, raising their own right hand in perfect synchronization. Their movements are perfectly aligned, officially launching the performance. This seamless transition from zoom to singing draws the audience’s attention away from the surrounding environment and fully into the performers’ actions and vocal delivery.

From 5 to 9 seconds, the stable collaborative rap section unfolds. With the zoom locked in place, both performers remain within the frame. Performer B withdraws their pointing hand and alternates between classic rap gestures—spreading their hands, flipping them over, and making expressive motions—while their body sways gently from side to side in rhythm, occasionally turning to interact with Performer A. Performer A keeps their eyes closed, focused on singing into the microphone, their head nodding subtly to the flow, while their hands perform alternating finger‑pointing and interlaced gestures. Their body leans slightly toward Performer B, creating a natural partnership where A emphasizes immersive vocal expression through restrained, precise movements, and B amplifies the rhythm with more outward‑directed gestures. This dynamic interplay ensures the visuals remain cohesive without becoming chaotic.

At 9 to 13 seconds, the scene shifts dramatically. The camera remains fixed on a medium close‑up. Performer B introduces dance‑like movement: crossing their arms in front of them, they pivot their torso to the right—away from the camera—gradually turning their back to it. Their arms swing gracefully in time with the beat, their feet shuffle slightly, revealing the purple patchwork design on their back, while their body undulates softly in sync with the rhythm. Meanwhile, Performer A maintains their stance facing the microphone, continuously holding the “gun‑finger” gesture formed by extending the index finger and thumb, their gaze firmly locked on the mic, completely unfazed by Performer B’s turn. This shift aligns perfectly with the eighth‑beat marker, injecting fresh dynamism into the scene while still preserving the steady rhythm established earlier. Performer A’s unwavering stability ensures the performance stays grounded and unhindered.

From 13 to 16 seconds, the flow takes another turn. Performer B pivots back to face Performer A and the camera, simultaneously extending their index fingers toward each other, then quickly forming a “shh” gesture with their fingers pressed to their lips—perfectly timed to coincide with the heavy beat marking the flow change. After lowering their hands, Performer B makes a light gesture brushing their forehead, almost like a salute, while Performer A spreads their hands wide in front of them. Both wear relaxed expressions, exuding effortless ease and confidence. These synchronized gestures further strengthen their rapport, pushing the emotional intensity of the segment to a minor climax.

Finally, from 16 to 20 seconds, the performance winds down. The camera slowly pulls back via optical zoom, revealing the full yellow platform and the black speaker in the lower left corner, returning to the symmetrical framing of the opening. Performer B raises their hand to point upward at the microphone overhead, then lowers it. Performer A lifts their hands in a gentle spreading motion, and together they make subtle adjustments to their positioning, facing each other and responding to the final drumbeats with slight body bounces. Performer B briefly sweeps their hands and clothes, looks up at Performer A, and finally stands still together beneath the microphone, bringing the piece to a graceful close. The pull‑back action synchronizes seamlessly with the concluding moments, completing the circular structure of the opening panoramic shot and allowing the rhythm to fade naturally from its peak back to calm.

Throughout the video, the cinematography remains deliberately restrained, eschewing unnecessary camera movements or cuts to prioritize the rap performance itself. Every gesture, turn, and interactive motion is precisely aligned with the beat, and both performers maintain a subtle knee‑bouncing baseline rhythm—a quintessential live‑rap physicality. Clear division of roles, tight coordination, and meticulous control over pacing combine to deliver a 20‑second performance that feels both fluid and memorable."

Hotel Lobby

Generate a 10-second hyperrealistic video with 4K, cinematic-quality texture, {camera 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, temperament, hair color, hairstyle, makeup, and outfit are 100% based on the characteristics depicted in @the image.
[10-Second Video]
This is a fast-paced, horror‑style short video that employs hard cuts and eye‑line matching throughout its editing. By rapidly alternating shot sizes and angles, it piles up suspenseful moments, culminating in an absurd twist involving spatial dislocation:
0:00–0:02 – Setting the stage
Shot: Medium shot, static camera positioned at eye level with no movement; shows a bright, everyday residential stairwell: white walls adorned with radiating metallic art pieces, a sheer white curtain on the right, wooden handrails with black brackets, and power outlets and switches visible on the wall.
Action: The protagonist ascends the stairs, initially resting their right hand on the rail while stepping up with their left leg. Suddenly, they look up toward the top of the staircase, slide their right hand along the rail to grip the front end tightly, bend their knees, instinctively shift their center of gravity backward, and display a wary, startled expression, their gaze locked firmly on the upper area.
0:02–0:02.8 – Introducing the first scare
Transition: Hard cut, matched by eye‑line editing (fully continuing the direction of the protagonist’s upward gaze).
Shot: Low-angle, upward‑facing fixed camera; reveals a bedroom scene: beneath the white ceiling, a sliding wardrobe with black frames and white doors reflects the TV on the wall and the bedroom door, while a terrifying figure stands atop the wardrobe.
Visuals & Action: A ghostly face—pale foundation paired with dark smoky eye makeup—wears a black hooded layered gauze robe, standing rigidly atop the wardrobe. Its expressionless face looks straight down at the camera, and the extreme upward angle dramatically amplifies the oppressive sense of being watched from above.
0:02.8–0:03.5 – Heightening fear
Transition: Hard cut, matched by eye‑line/reaction.
Shot: Medium‑closeup, static camera at eye level, focusing on the protagonist’s upper body as they return to the stairwell scene.
Action: The man continues gripping the handrail tightly, eyes wide open, mouth slightly agape, facial muscles tense, maintaining his upward gaze. His body remains frozen, utterly stunned by what he sees above.
0:03.5–0:05.5 – Building tension
Transition: Hard cut, returning to the scene where the spirit resides.
Shot: Low‑angle, upward‑facing fixed camera, leveraging the character’s forward lean to create a perspective effect that pushes the frame closer.
Action: The spirit slowly leans forward from its rigid stance, extending a pale right hand downward toward the camera (in the direction of the stairs). As the hand grows larger in the frame, its fingertips point toward the lower portion of the shot, while the ghostly face keeps its gaze fixed below. This gradual motion intensifies the feeling of impending capture and mounting pressure.
0:05.5–0:06.8 – A pause in rhythm, introducing new suspense
Transition: Hard cut, back to the stairwell.
Shot: Medium shot, static camera at eye level.
Action: The protagonist maintains their grip on the handrail but suddenly glances toward the room door where the spirit resides (on the right side of the frame), their expression shifting from terror to puzzled bewilderment—as if hearing a sound. Their attention shifts away from the spirit above, and they subtly pivot their body to the right, slightly moving their weight forward.
0:06.8–0:08.5 – Unveiling the second scare
Transition: Hard cut, matched by perspective (a sideways view).
Shot: Close-up of the floor, extremely low‑angle俯拍 (close to the tiled surface), static camera; reveals the bedroom’s white tiled floor, with a yellow floral bedspread draped over the bed on the right.
Visuals: Two very thin black tentacles protrude from under the bed, connecting to a small dark patch beneath that resembles a giant insect. These tentacles tremble ever so slightly, hinting at something hidden under the bed—a classic horror motif that heightens tension.
0:08.5–0:10 – The twist finale
Transition: Hard cut, returning to the overhead view of the wardrobe.
Shot: Low‑angle, upward‑facing fixed camera.
Visuals & Action: The black-robed spirit still stands atop the wardrobe, while the protagonist—who was previously on the stairs—now finds themselves perched on the edge of the wardrobe to the spirit’s right. With their left hand pressed against their chest, looking dazed and shaken, the protagonist clutches the spirit’s hand tightly. Together, they peer downward, completing the chilling twist with an absurd scene that defies conventional spatial logic.
Overall Cinematography and Narrative Style
Cinematic Language: Throughout, there are no complex camera movements. Instead, static medium shots convey character reactions, low‑angle upward shots amplify the spirit’s oppressive presence, and extreme close-ups create an atmosphere of unknown dread—all alternating swiftly to stir audience emotions.
Editing Logic: All transitions rely on hard cuts, seamlessly linking scenes according to the protagonist’s line of sight. The pacing follows the standard short‑video narrative structure of “everyday setup → introduction of scares → emotional buildup → momentary pause → secondary scare → climactic twist,” delivering three distinct emotional peaks within just ten seconds.

Haunted

Generate a 10-second hyperrealistic video with 4K HD cinematic-quality realism. {Camera shooting pixel parameters: 1.6}, natural image noise level: 3.0, {Realism parameter: 1.6}. The lighting and shadow transitions are seamless, and the human body proportions are perfectly accurate. The protagonist’s facial features, gender, temperament, hair color, hairstyle, makeup, and clothing style are 100% based on the reference @image.  

[Video: 10 seconds]  
The entire video is shot using a fixed, forward-facing medium-shot angle from inside the car, with the camera positioned behind the car’s dashboard, directly facing both the driver’s seat and the passenger seat. The framing remains stable, capturing only the upper body of the characters throughout. There are no tracking shots, zooms, pans, or movements during the entire sequence. The core narrative employs a parallel montage technique with rapid flash cuts, alternating between two timelines at a pace of roughly one second per shot. The two timelines are perfectly aligned in terms of camera position, relying solely on lighting and character states to create stark contrasts:  

**Real-time timeline:** A cool-toned night scene. The female lead sits alone in the driver’s seat of a luxury sedan; the passenger seat is empty. Outside the car, neon lights flicker softly across the urban streetscape.  

**Memory timeline:** A warm-toned, dusk-time backlight scene. Both protagonists share the driver’s seat of an old family sedan—the female lead occupies the driver’s seat, while the male lead sits in the passenger seat. Outside the car, the setting sun casts golden hues over a suburban meadow.  

The final shot returns to the real-time timeline, completing the emotional loop.  

Action timing for both protagonists (broken down by time segments):  
0–1 second: The opening of real-world emotions—In the real world, the female lead rests one hand on the steering wheel and uses the other to cover her mouth. When she lowers her hand, tears streak her face, and her expression is filled with sorrow and detachment.  
1–2 seconds: A sweet transition into memory—Cut to the dusk-time memory scene, a side medium shot. The female lead sits in the driver’s seat, hands gripping the steering wheel, turning her head to smile at the male lead in the passenger seat. The male lead turns his head and gently kisses her cheek. Both protagonists exude relaxed, tender energy.  
2–3 seconds: An outburst of real-world emotions—Back in the real world, the female lead clenches both hands around the steering wheel, sobbing silently, releasing her inner pain.  
3–4 seconds: Memory interaction—Back in the memory scene, the female lead turns her head slightly in the driver’s seat, holding the steering wheel with both hands. She turns her head to smile at the male lead, who responds with a gentle, loving gaze. The atmosphere is light and intimate.  
4–5 seconds: The real-world emotional climax—Back in the real world, the female lead grabs a handful of banknotes from the passenger seat and hurls them onto the passenger seat, scattering them everywhere.  
5–6 seconds: Joyful memory interaction—Back in the memory scene, the female lead sits atop the male lead in the old sedan, sharing a sweet kiss. No banknotes from the real world appear in this memory scene.  
6–8 seconds: Real-world breakdown and accusation—Back in the real world, the female lead’s eyes are red-rimmed, and she points toward the empty passenger seat piled high with banknotes. Her anger and despair are palpable, her facial expression distorted with intense emotion.  
8–9 seconds: A tender, frozen moment in memory—Back in the memory scene, the female lead returns to the driver’s seat, holding the steering wheel with one hand. The male lead takes her other hand in his, gently kissing the back of her hand. Both close their eyes, radiating tender affection.  
9–10 seconds: The real-world finale—a lonely, desolate ending—Back in the nighttime luxury car, the female lead holds the steering wheel with one hand and raises the other to wipe away tears from the corner of her eye. She continues to grip the steering wheel with both hands, her expression vacant and forlorn. The passenger seat remains empty. The shot fades out.  

Note: There are no specific requirements for clothing. In the real-time timeline, the female lead wears formal attire; in the memory timeline, both protagonists wear casual, everyday clothes. This difference in wardrobe accentuates the contrast between the two timelines. Throughout the video, the characters remain silent, with only background music providing auditory accompaniment.

Regret

"Generate a hyperrealistic 15-second video with 4K HD cinematic-quality texture. {Camera shooting pixel parameters: 1.6}, natural image noise level: 3.0, {Realism parameter: 1.6}. The light and shadow transitions are seamless, and the human body proportions are perfectly accurate. The protagonist’s facial features, gender, demeanor, hair color, hairstyle, and makeup are 100% based on the facial characteristics depicted in the reference @image.

[Video: 15 Seconds]

This 15-second short film unfolds in an open, overcast meadow on a suburban field, featuring a narrative filled with twists and suspenseful tension. The protagonist, dressed in a light-gray plaid suit paired with a gray-green tie, walks toward the edge of a freshly dug pit. From within the pit, several hands—of varying skin tones—reach upward, fingers spread wide, continuously reaching out in a gesture of help.

The protagonist lowers his gaze to the pit at his feet. First, he reaches into his suit pocket, pulls out a wallet, and carefully extracts a few banknotes, handing them down into the pit. As soon as the hands inside the pit grasp the bills, the protagonist crouches down, leans over the edge of the pit, and firmly grabs one of the hands that has reached up, exerting force to pull it upward, attempting to lift the person out of the pit. Unexpectedly, the person inside the pit suddenly tightens their grip on the protagonist’s arm, pulling him forcefully backward. The protagonist’s body lurches forward, nearly losing balance and falling into the pit. Instantly, he stabilizes his stance, sits down on the edge of the pit, and with all his strength yanks his arm back, straightening up once again and regaining his footing by the pit’s edge.

Lowering his head, the protagonist silently watches the hand still reaching upward. After a brief pause, he no longer makes any attempt to offer assistance. Instead, he turns around, takes a step, and walks straight across the meadow into the distance. Meanwhile, the hand inside the pit slowly sinks back into the depths of the pit.

Throughout the entire sequence, the camera remains fixed in a medium shot, without any zooms, pans, or changes in shot composition. The frame consistently captures both the pit and the protagonist standing by its edge in full view. The cool-toned, overcast natural light evenly bathes the green grass and the characters, while distant trees serve as a softly blurred background. The camera remains steady and motionless, fully documenting the entire process—from the initial rescue attempt, through being pulled back, to the protagonist’s final turn and departure. There is no unnecessary camera intervention; instead, the event’s dramatic shift is presented from an objective, restrained perspective."

Deception

Use the uploaded reference video as the exact motion, choreography, timing, composition, and camera reference.
MAIN CHARACTER:
Replace ONLY the single central main character with the person shown in Character Reference 1.
The central main character must exactly match Character Reference 1 throughout the entire video.
Preserve the exact facial identity, facial features, hairstyle, skin tone, body proportions, clothing, and overall appearance of Character Reference 1.
The central character must remain clearly distinguishable from all surrounding characters at all times.
SURROUNDING CHARACTERS:
Replace ALL surrounding background people with the characters shown in Character Reference 2.
Every surrounding person must consistently match Character Reference 2 in clothing, headwear, mask, colors, body proportions, and overall appearance.
Do NOT apply Character Reference 2 to the central main character.
Do NOT apply Character Reference 1 to any surrounding character.
ACTION AND CHOREOGRAPHY:
Reproduce the exact actions, body movements, choreography, poses, timing, rhythm, and synchronization from the original reference video.
The surrounding crowd must perform the same synchronized group movements and synchronized deep bowing actions shown in the reference video.
The central main character must perform only the movements of the original central character.
Do not invent new gestures, movements, interactions, or choreography.
IDENTITY LOCK:
Keep the identity of the central character completely consistent from the first frame to the final frame.
Do not change, morph, merge, distort, duplicate, or swap the central character's face.
Do not replace the central character with any background character.
Keep all background characters visually consistent throughout the entire sequence.
CROWD CONSISTENCY:
Preserve the original number, placement, spacing, formation, depth, and relative positions of the surrounding people.
Do not add or remove people.
Do not randomly move people between positions.
Do not merge bodies or duplicate characters.
MOTION LOCK:
Strictly preserve the original video's motion.
Preserve the exact body movements, movement direction, choreography, synchronization, pacing, rhythm, and action timing.
CAMERA AND FRAMING:
Keep the original camera angle, camera movement, camera position, focal length, perspective, depth of field, framing, composition, shot scale, and shot duration unchanged.
Follow every original camera movement exactly.
Do not introduce any new camera movement.
Do not re-frame, crop, zoom, rotate, tilt, or alter the original composition unless that exact movement exists in the reference video.
SCENE:
Preserve the original spatial arrangement, crowd formation, foreground-background relationship, and scene geometry.
Maintain natural lighting, realistic shadows, realistic human anatomy, and photorealistic textures.
TEMPORAL CONSISTENCY:
Maintain strong frame-to-frame consistency.
No flickering.
No face changes.
No clothing changes.
No disappearing characters.
No duplicated people.
No warped hands or limbs.
No sudden changes in body proportions.
FINAL REQUIREMENT:
The final video must follow the reference video's original motion, choreography, timing, crowd formation, camera work, and editing as closely as possible, while changing only the visual identities specified by Character Reference 1 and Character Reference 2.

Chosen

Generate a 12-second hyperrealistic video with 4K high-definition, cinematic-quality texture. {Camera shooting pixel parameters: 1.6}, natural image noise level: 3.0, {Realism parameter: 1.6}. The light and shadow transitions are seamless, and the human body proportions are perfectly accurate. The protagonist’s clothing, facial features, skin tone, and hairstyle are 100% based on the reference images provided.

[First Frame]

Low-angle upward shot, close-up camera position, with focus centered on the protagonist’s feet: The protagonist is wearing bright, glossy red leather Western-style tall pointed-toe boots with naturally worn creases along the boot shaft. The boots feature thin high heels, and the protagonist is taking a step forward, with the toe of the boot about to land on the white crosswalk line of a city zebra crossing, partially revealing the lower calf skin. The background is a blurred, daytime urban street scene, where you can vaguely make out gray traffic signal poles, sidewalk edges, red vehicles parked along the curb, and glass-clad high-rise buildings. The ground is a dark gray asphalt surface sprinkled with fine white particles. The white paint of the zebra crossing shows slight signs of wear. Overall, the scene captures a natural street atmosphere under daylight, with the blurred background accentuating the red boots as the main subject.

[Video: 12 Seconds] Continuous shot, no cuts or lens transitions.

0-1s: Low-angle upward shot focusing on the protagonist’s feet—specifically, the bright, glossy red leather Western-style tall pointed-toe high-heeled boots. The boot shaft has natural creases from wear, and the protagonist is walking forward along the zebra crossing, with the toe of the boot about to touch the white crosswalk line that shows slight signs of wear. The background is blurred, revealing gray traffic signal poles, red vehicles parked along the curb, glass-clad high-rise buildings, and a dark gray asphalt road surface dotted with fine white reflective particles.

1-3s: The camera smoothly pans upward, transitioning naturally from the close-up of the feet to a medium shot, gradually revealing the protagonist’s full outfit. The protagonist’s expression is relaxed, lips slightly parted, and their posture is loose and natural. In the background, you can see a traffic signal with a red “No Pedestrians” sign lit up, as well as street vehicles and buildings.

3-5s: On the right side of the frame, a passing pedestrian’s partial body appears. The protagonist’s shoulder collides with the pedestrian’s shoulder, causing the sunglasses she’s wearing, the mobile phone she’s holding near her ear, and the ice drink she’s clutching in her hand to fly into the air simultaneously. The protagonist instantly spreads both hands wide to the sides, instinctively reaching out to catch the objects. Her expression shifts from relaxed to one of shock and surprise; her mouth is slightly open, and her eyes fixate on the objects flying away.

5-7s: The camera zooms forward, tracking the floating objects in mid-air at a high frame rate and slow motion. The focus first locks onto the sunglasses in the air, showing the lenses and then the temple on one side. Behind it, the blurred floating ice drink begins to spill cola liquid from its cup rim. The protagonist remains blurred in the background, keeping her hands spread wide in a shallow depth-of-field effect that highlights the sunglasses as the main subject.

7-9s: At a high frame rate and slow motion, the ice drink cup slowly floats toward the camera. The camera’s focus switches to the ice drink cup in mid-air, following its gentle movement. Inside the transparent plastic cup, the dark-colored cola with ice cubes shakes violently due to inertia, causing the ice cubes to collide within the cup. Dark-brown cola liquid splashes out of the cup rim in varying-sized droplets, while tiny water droplets condense on the cup’s walls. The protagonist remains completely blurred in the background, with the focus entirely locked on the splashing cola and the ice-cup object, vividly capturing the dynamic motion of the droplets suspended in the air.

9-11s: At a high frame rate and slow motion, the flipped-open mobile phone slowly floats toward the camera. The camera’s focus switches to the flipped-open mobile phone in mid-air, following its trajectory as it flips over. The silver-colored flip phone is open, with its screen lit up displaying the interface. As it slowly flips and falls through the air, you can faintly see the protagonist’s outstretched hand reaching for the phone in the background—but she never quite manages to grab it. The shallow depth-of-field effect emphasizes the floating, flipping phone as the main subject.

11-12s: The camera tracks downward as the phone falls, eventually settling at ground level. The phone automatically closes itself in mid-air and lands on the dark gray asphalt road surface. Upon contact with the ground, the phone’s screen remains lit. In the background, you can see the blurred legs of passing pedestrians moving in the distance. The focus is firmly fixed on the fallen phone, with the granular texture of the asphalt clearly visible. The pedestrians in the background appear as blurred, moving silhouettes.

Urban Slow Motion

Generate a hyperrealistic 12-second video with 4K HD cinematic-quality realism. Camera shooting pixel parameters: 1.6; natural image noise: 3.0; realism parameter: 1.6. The lighting and shadow transitions are seamless, and the human body proportions are perfectly accurate. The protagonist’s facial features, skin tone, hairstyle, and overall demeanor remain 100% consistent with those in Reference Image 1.

The core metaphor of this video is “self-redemption.” The rescuer and the person who has fallen into the water represent two different states of the same protagonist. Only two states of the same protagonist are allowed to appear in the footage—no second, independent character should be generated. Both states must share the same face, the same facial structure, the same skin tone, the same hairstyle, the same body type, the same gender, and the same overall demeanor. They should be distinguished solely by differences in wetness, positioning, clothing condition, and emotional expression. Absolutely no third parties, different facial features, different genders, or different identities should appear.

The clothing gender requirements must strictly follow Reference Image 1:

If the protagonist in Reference Image 1 is female, in the drowning state she wears a long, off-white dress made of soft fabric. After being completely soaked by seawater, the dress naturally conforms to her body. As it absorbs water, the hem becomes heavy, droops gracefully, and is gently swayed by the sea breeze, maintaining an elegant, restrained, and non-exposing visual effect.

If the protagonist in Reference Image 1 is male, in the drowning state he wears a long-sleeved, off-white or light-gray shirt suitable for men, made of linen or lightweight cotton. After being soaked by seawater, the shirt naturally clings to his body. He wears dark-gray or black trousers, whose hems are dampened by the seawater. The overall look is simple, neat, and in line with everyday men’s attire. The male version should not include dresses, skirts, or feminine clothing.

The clothing in the redemption state must also match the protagonist’s gender: For females, wear a long black coat and black boots; for males, wear a long black waterproof windbreaker or black overcoat, dark-colored trousers, and black ankle boots. Both states must belong to the same person—gender changes are not permitted.

I. First Frame Description

This is a vertical-format, cold-toned, cinematic narrative shot. The camera angle is a close-up, level view. The lighting is cool gray diffuse natural light under an overcast sky, creating an overall atmosphere of oppression and despair, filled with the feeling of helplessness after struggling to stay afloat.

The scene is set on a slippery, black rocky cliff by the sea. The rock surface is wet from seawater, reflecting natural water highlights. The background is a dull, gloomy overcast sky. In the frame, the drowning protagonist’s body is fully soaked by seawater. Her wet hair clings to her face and neck, tears mixed with seawater streaming down her cheeks. She’s breathing heavily, exhausted and on the verge of collapse.

If the protagonist is female: She wears a long, off-white dress that’s completely soaked but remains natural and opaque. The hem becomes heavy and droopy as it absorbs water. She stands barefoot or hangs precariously at the edge of the slippery rocks.

If the protagonist is male: He wears a long-sleeved, off-white or light-gray shirt that’s soaked through, paired with dark-gray or black trousers. The hem of the shirt and the trouser legs are dampened by seawater. He’s barefoot or wearing simple shoes and socks that have been wetted by the sea.

In the drowning state, the protagonist’s hands grip tightly onto the edge of the rocks, her body partially hanging below the ledge. Her profile faces the camera, her shoulders trembling slightly from exhaustion and sobbing. Above the frame, leave space between the rock edge and the gray sky, which will gradually reveal the redeemed state of the same protagonist later—but do not introduce any additional characters.

II. Shot Breakdown and Detailed Analysis

The entire video consists of a single, continuous shot, narrated through a slow zoom-out movement without any hard cuts. It’s divided according to emotional and action progression as follows:

0–3 seconds: The stage of desperate struggle

The mood is cold and cinematic, with high-contrast cool tones. The reflections of wet fabric and wet rocks feel realistic and natural. The waves move softly, and the overall atmosphere is heavy and oppressive.

The drowning protagonist’s hands firmly grip the edge of the rocks, her body still hanging below the ledge. Her forehead rests against the rock as she breathes heavily, her shoulders trembling slightly from exhaustion and crying. Her wet hair and clothes are gently stirred by the sea breeze. If she’s female, the soaked-off-white long dress and its heavy hem sway slowly in the breeze. If she’s male, the soaked-off-white or light-gray long-sleeved shirt and dark-colored trousers are moved by the wind.

The camera stays fixed on a close-up, with extremely subtle natural breathing-like vibrations. The focus is on the protagonist’s upper body, face, and hands. Her brows are tightly furrowed, her eyes half-open, her face streaked with tears. Her expression is painful and exhausted, on the verge of giving up.

3–6 seconds: The stage when hope begins to emerge

The lighting is cool and soft, the sea breeze feels natural, and the mood gradually shifts from oppression toward a turning point. The camera starts to slowly zoom out, revealing the full figure of the “redeemed protagonist” above the rocks.

The redeemed protagonist must have exactly the same face, facial features, skin tone, hairstyle, body type, gender, and overall demeanor as the drowning protagonist—only their hair and clothes are drier, their posture more stable, and their emotions calmer. The female version wears a long black coat and black boots; the male version wears a long black waterproof windbreaker or black overcoat, dark-colored trousers, and black ankle boots.

The redeemed protagonist’s hair is blown by the sea breeze, and they’re looking down at their own submerged self, slowly reaching out with one hand holding a silver ring. The drowning protagonist slowly lifts her head and looks upward, releasing one hand and trembling as she reaches out to grasp that hand. The eyes of both protagonists must form a before-and-after comparison of the same person, as if meeting their future self.

The camera smoothly and slowly zooms out, transitioning from a close-up to a medium shot, gradually bringing the redeemed protagonist above the rocks into the frame. This clearly shows that these are two states of the same person, not two strangers.

6–9 seconds: The stage of vigorous rescue

The motion carries slight motion blur, emphasizing the sense of effort. In the cold-toned atmosphere, a healing sensation begins to emerge. The two hands grip each other tightly. The redeemed protagonist exerts force to pull upward, while the drowning protagonist clenches her teeth, lifts her bare foot or a seawater-soaked foot to brace against the side of the rock, using all her strength to climb upward. Finally, she steps onto the rocky platform, standing completely soaked in front of her redeemed self.

The camera rises slightly along with the climbing motion, maintaining a medium-shot composition. The focus remains on the hands clasped together and the climbing action. The drowning protagonist’s expression shifts from disbelief to intense concentration during the exertion. At the moment she steps onto the shore, her gaze relaxes, and all her defenses gradually fall away.

If she’s female, the soaked-off-white long dress drags naturally, gets wet, and clings to her legs as she climbs. If she’s male, the soaked long-sleeved shirt, dark-colored trousers, and lower garment hems swing naturally with the climbing motion. The change in clothing is only used to distinguish between the drowning and redeemed states—it doesn’t alter the protagonist’s identity.

9–12 seconds: The stage of embracing redemption

The atmosphere is open and cool, with the backdrop of waves highlighting the emotional weight. The healing sensation is strong. The redeemed protagonist opens her arms wide, pulling her soaked self tightly into her embrace. One hand gently strokes the back and hair of the drowning protagonist, offering comfort.

The drowning protagonist buries her face in the other’s shoulder, closes her eyes and sheds more tears. Her hands wrap around the other’s back, completely releasing her emotions. The hair and clothes of both protagonists are stirred by the sea breeze: In the female version, the wet dress, black coat, and hem move dynamically; in the male version, the wet long-sleeved shirt, dark-colored trousers, and black windbreaker hem move dynamically.

The two protagonists stand quietly embracing on the rocks. The camera maintains a medium shot and slightly zooms out, fully capturing the deep blue sea behind them, where white waves surge. The figures appear soft yet powerful against the vast sea and sky. In the embrace, the drowning protagonist gradually relaxes from trembling and sobbing, while the redeemed protagonist’s gaze becomes calm and gentle, steadily supporting the other’s emotions.

The ending must clearly convey: Embracing yourself is embracing the self that has emerged. The rescuer isn’t someone else—it’s the self that has risen again after experiencing pain.

{First-frame composition weight: Slippery black rocks by the sea, drowning protagonist, the redeemed protagonist gradually revealed above, single-person dual-state self-redemption metaphor: 1.8}  
{Clothing condition weight: Female uses a long off-white dress; male uses a long off-white or light-gray soaked long-sleeved shirt paired with dark-colored trousers. Clothing must match the reference image protagonist’s gender: 1.8}  
{Camera movement weight: Continuous single shot, slow zoom-out, slight natural shake, no hard cuts, clear emotional progression: 1.7}  
{Action emphasis weight: Self-pulling, same protagonist in dual states, climbing onto the shore, embracing oneself, sense of redemption: 1.8}  
{Background/style weight: Overcast cold gray, slippery rocks, deep blue sea, dull gray sky, cinematic-realistic mood: 1.7}  
{Prohibited elements weight: No second independent character references, no third parties, no gender changes, no men wearing dresses, no mismatched feminine/masculine clothing, no indoor scenes, no bright sunny days, no cartoonish style, no exaggerated effects, no incorrect proportions, no abrupt cuts, no character clipping, no limb deformation, no extra limbs, no excessive skin smoothing, no transparent or revealing clothing: 1.6}  
{The protagonist’s main facial features, skin tone, hairstyle, body type, gender, and overall demeanor in the reference image remain unchanged}  
{Realism: 1.6}  
Human body proportions are perfectly accurate.

Falling

Generate a 15-second video with hyperrealistic visuals, 4K HD cinematic-quality texture, {camera shooting pixel parameters: 1.6}, natural image noise level: 3.0, {realism parameter: 1.6}, seamless light and shadow transitions, and perfectly accurate human body proportions.

The protagonist’s facial features, gender, temperament, skin tone, age characteristics, hair color, hairstyle, makeup, body type, and identity are 100% based on the reference image of the protagonist.

The animal’s species, appearance, coat color, body shape, facial features, normal eye color, fur, feathers, scales, body proportions, and identity are 100% based on the reference image of the animal. The animal is not fixed to any specific species; its species is not replaced, and its original body structure and physical characteristics remain unchanged.

[Video—15 seconds]

The entire video is shot using fixed close-up, medium, and extreme close-up shots, with no zooming, panning, or rotating camera movements. Hard cuts are used as the primary transition method, with only one soft, warm-light overlay fade used when entering flashback sequences. Once the flashback ends and the scene returns to reality, clear hard cuts or motion-matching transitions are employed to depict the zombie-like animal lunging at the protagonist’s neck and biting them, subsequently triggering infection and mutation due to the bite wound.

The overall composition follows a cinematic widescreen format with black borders at the top and bottom. Only the protagonist and the animal from the uploaded image appear in the frame; no additional characters, animals, background passersby, silhouettes of people, or silhouettes of animals are added.

The setting is a dark, oppressive outdoor ruin environment with dusty cement ground. The background is completely blurred, punctuated by a few warm-yellow bokeh highlights. The overall color palette is low-saturation, realistic, with a balanced mix of cool and warm tones, featuring slight film grain, shallow depth of field, and realistic environmental dust particles.

The protagonist wears a khaki tactical vest, black shoulder straps, and a light gray-green combat suit. The animal is outfitted with camouflage tactical gear that matches its body structure and fits its species, while retaining the animal’s true species characteristics, body proportions, fur, feathers, scales, or other surface details from the reference image.

The core emotional arc is as follows:

Sad crying → Tense confrontation → Appearance of the zombie-like animal → Struggling in pain → Close-up of the zombie-like animal → Warm daily memory → End of the memory → In reality, the animal attacks the protagonist’s neck → Bite wound leads to infection → Protagonist undergoes mutation.

0–1s: Wounded Crying Segment

The opening shot is a close-up of the protagonist’s face, occupying over 70% of the frame. The protagonist maintains the hair color, hairstyle, facial structure, skin tone, and identity characteristics from the reference image.

The protagonist has a scrape and slight bleeding on the left side of the forehead, a small scratch on the left cheek, residual bloodstains on the right corner of the mouth, and clear tear tracks covering the face. Transparent teardrops slowly slide down from the lower eyelid, and the skin appears moist and glossy due to intense emotion.

The protagonist’s eyebrows are slightly furrowed, their gaze sorrowfully downward, lips slightly parted and trembling gently, in a state of suppressed, broken crying. Soft sidelighting creates natural light and shadow transitions across the face, highlighting the wounds, tears, and authentic skin texture.

The background is a completely blurred, dark-toned outdoor ruin with a few warm-yellow bokeh highlights. The shot is a fixed close-up, with the protagonist mostly still, only the tears slowly sliding down at a restrained, subdued pace.

1–3s: Appearance of the Zombie-Like Animal Segment

A hard cut transitions to an over-the-shoulder medium shot behind the protagonist. In the foreground is the holographic sight of an assault rifle, sharply blurred; in the middle ground is the animal from the reference image.

The animal wears or carries camouflage tactical gear that fits its body structure, standing or sitting upright in a natural pose consistent with its species, resting on the dusty cement ground.

As soon as the animal appears in the frame, its eyes already display abnormal zombie-like characteristics: they’re grayish-white, cloudy, lacking luster, with blurry or nearly invisible irises and pupils, giving off an unnatural milky-gray hue and a deathly stillness.

From the moment it first appears, the animal maintains this zombie-like eye state. It’s forbidden for the animal to start with normal eyes or show a gradual change from normal to milky-white.

The animal faces the camera directly, appearing unnaturally calm, serious, and vacant, devoid of the lively expressions commonly seen in ordinary animals. The animal slowly turns its head to one side, while its tail, wings, tentacles, or body remains naturally stable—its movements must match the real body structure of the animal in the reference image.

If the animal in the reference image isn’t a quadruped, its movements will be adapted according to its actual body structure, locomotion style, and posture, without forcing feline, canine, or other fixed-species motions.

The background is a backlit, abandoned outdoor street, where warm golden sunlight creates soft Tyndall effects, and fine dust particles float in the air. The shot is a fixed over-the-shoulder medium, with the foreground firearm slightly blurred, the focus locked on the animal. The overall atmosphere is tense, eerie, and unsettling—a cinematic, hyperrealistic feel, captured in 4K HD.

3–6s: Painful Aiming Segment

A hard cut returns to a frontal close-up of the protagonist. The foreground retains the holographic sight of the assault rifle; the metal edges of the gun are visible but slightly blurred, with the focus firmly on the protagonist’s face.

The protagonist maintains the gun-aiming stance, shoulders trembling slightly, eyebrows tightly furrowed, and eyes showing intense pain and struggle through the scope. Tears continuously well up in the eyes, streaming down the cheeks, lips trembling and slowly parting—as if emotions have already broken down yet the protagonist can’t pull the trigger.

Soft sidelighting accentuates the tear tracks, scars, and moist texture of the protagonist’s skin, against a dark-blurred outdoor background. The shot is a fixed close-up, with no extra camera movement, slow and restrained pacing, amplifying the protagonist’s unbearable emotional agony.

6–8s: Close-Up of the Zombie-Like Animal’s Face Segment

A hard cut transitions to an extreme close-up of the animal’s face. This segment focuses solely on the animal’s face, eyes, and local body details—not on new infection processes or changes in eye color.

The animal maintains the zombie-like eye state it first displayed during the 1–3 second mark: both eyes remain grayish-white, cloudy, and lifeless, with blurry or absent irises and pupils, and an otherworldly milky-gray hue inside the eyes.

The animal stays calm, stiff, and vacant, slowly performing a natural eye-opening or blinking motion. Its fur, feathers, scales, ears, tentacles, tail, or other species-specific details exhibit extremely subtle movement.

The animal doesn’t change species, doesn’t alter its body structure, doesn’t adopt human-like expressions, and doesn’t undergo secondary mutations. This segment is exclusively focused on the zombie-like animal’s face and eyes.

The background is a blurred, warm-light outdoor setting, with fine dust particles floating in the air. The shot is an extreme close-up, with the focus entirely on the animal’s zombie-like eyes and facial details. The atmosphere is eerie, calm, and oppressive—a hyperrealistic horror movie vibe.

8–10.2s: Warm Daily Memory Segment

The scene transitions via a soft, warm-light overlay fade into a pre-infection memory sequence.

This segment is a completely separate, warm, everyday memory—showcasing only peaceful, intimate, and happy moments between the protagonist and the animal. There’s no aggression, biting, chasing, struggling, injury, bleeding, mutation, or horror elements in this memory.

The indoor space is illuminated by a warm golden sunset, with light filtering through sheer curtains, creating soft halos around the outlines of the protagonist and the animal.

The protagonist wears a pink turtleneck sweater, holding the animal in their arms and gently stroking its back or body. The protagonist closes their eyes and smiles happily, tilting their head back slightly, completely relaxed and blissful.

In the memory, the animal must remain healthy, normal, natural, and affectionate—its eyes return to normal color and natural luster, showing no zombie-like eyes, no milky-white eyes, no cloudy, lifeless gaze, and no signs of wounds or infection.

The animal can perform warm, everyday actions appropriate to its real species—for example, approaching the protagonist, rubbing against their face, snuggling, licking their cheek, quietly lying in their arms, gently lifting its head toward the protagonist, or naturally staying close.

If the animal lacks licking behavior, it’ll instead approach, rub against, snuggle, or engage in other affectionate actions fitting its real nature—without altering its natural habits.

The background includes cozy home elements like wooden cabinets, picture frames, and lamps—but the depth of field is shallowly blurred, avoiding obstruction of the protagonist and animal. The shot is a fixed, warm-toned medium, with a gentle, soothing rhythm. The overall color palette is warm gold, low contrast, and healing, creating a stark contrast with the earlier dark, oppressive reality scenes.

This segment must exclusively portray warm, everyday moments—no aggressive behaviors are allowed. It’s forbidden for the animal to suddenly attack the protagonist in the memory, and it’s forbidden for the memory scene to directly transition into an attack scene.

10.2–10.5s: Leaving the Memory and Returning to Reality Segment

The warm memory must end here completely. The protagonist and animal’s affectionate daily actions cease, and the golden sunset, indoor home environment, and warm lighting quickly fade away.

The scene transitions via a clear hard cut, brief foreground occlusion, or motion-matching transition, leaving the memory and returning to the dark, oppressive ruin environment.

It’s forbidden to complete the bite attack within the indoor memory scene, and it’s forbidden for the memory scene to directly turn into an attack scene. The two time periods must be clearly distinguished:

First, the warm, normal memory;

Then, the memory ends;

Next, the scene cuts back to the real-world ruin;

Finally, the zombie-like animal attacks the protagonist.

After returning to reality, the protagonist remains wounded, crying, and holding the gun in a defensive stance. The animal also reverts to its zombie-like state in the real-world setting, with eyes once again displaying the grayish-white, cloudy, and lifeless zombie-like appearance.

10.5–11.5s: Real-World Animal Attack on the Protagonist’s Neck Segment

Back in reality, the protagonist is still in the dark ruin environment, with tear tracks and wounds on their face, body maintaining a defensive posture after being attacked by the animal.

At this point, the animal is no longer the healthy, gentle creature from the memory—it’s now a zombie-like animal in the real world. The animal abruptly stops its still pose, its body instantly tensing, and its grayish-white, cloudy eyes lock onto the protagonist’s neck.

Then, the animal suddenly lunges from the real-world setting toward the protagonist, rapidly attacking their neck.

It’s crucial to emphasize:

The animal launches its attack only after leaving the warm memory and returning to the real-world ruin.

There’s absolutely no biting or harm occurring during the memory itself. It’s only after returning to reality that the zombie-like animal attacks the protagonist’s neck.

The animal uses its real mouth structure appropriate to its species, delivering a sudden bite or aggressive contact to the protagonist’s neck. If the animal has teeth, the bite is realistic yet restrained; if it has a beak, mouthparts, or other attack structures, the neck attack is portrayed according to its real body structure.

As the animal bites the protagonist’s neck, the protagonist shifts from a brief moment of daze and sadness to panic, pain, and struggle, instinctively recoiling backward.

The protagonist’s neck shows a clear bite wound or attack mark corresponding to the animal’s attack position. The wound displays only minimal blood and obvious damage, without excessive gore, tearing, dismemberment, or exaggerated spurting of blood.

The full causal relationship must be clearly shown:

The warm memory ends;

The scene cuts back to reality;

The zombie-like animal appears;

The animal attacks the protagonist’s neck;

A bite wound forms on the protagonist’s neck;

The bite wound triggers infection and mutation.

The animal’s head or body may briefly block the foreground of the shot, using foreground occlusion to complete the motion-matching hard cut—but the attack action must remain fully visible and understandable. Viewers must clearly see that the animal has bitten the protagonist’s neck.

It’s forbidden for the infection to appear before the animal’s attack, forbidden for the neck wound to appear before the attack, and forbidden for the wound to appear first and then be supplemented by the animal’s attack.

11.5–15s: Bite Wound Triggers Infection and Mutation Segment

A motion-matching hard cut returns to the dark ruin environment in reality. The protagonist lies on the dust-covered ground, maintaining the same posture after the animal’s attack, with the focus centered on the protagonist’s face, neck, and the recently bitten area.

The protagonist’s neck retains the bite wound inflicted by the animal in the previous shot—the wound’s position, direction, and spatial relationship must exactly match the animal’s attack location, ensuring continuity of action and space.

The mutation must begin precisely from the site of the animal’s bite, never appearing out of thin air or before the animal’s attack.

The protagonist first shows brief pain, panic, and incomprehension, followed by black, branch-like vascular patterns slowly spreading from the bite wound on the neck, gradually expanding to the neck, jaw, and entire face.

The protagonist’s skin slowly shifts from normal skin tone to a cold, grayish-blue hue, and their facial expression gradually becomes stiff, numb, and unfocused, starting from pain and fear.

The protagonist’s eyeballs slowly roll upward, the pupils gradually disappear, eventually turning into the milky-white, pupil-less state identical to the animal’s in reality.

A final tear still clings to the corner of the eye, suspended on the side of the face, creating a scene of combined sadness, infection, and despair.

The warm sidelight gradually gives way to a cold, grayish-blue tone. The shot is a fixed close-up, with no zooming or panning, slow and restrained movements, emphasizing the process of the bite wound, vascular spread, skin color change, eye whitening, and facial stiffness.

The ending freezes on the protagonist’s milky-white eyes, the neck bite wound, and the tear hanging on their face, clearly illustrating that the protagonist underwent infection and mutation precisely because they were bitten by the zombie-like animal upon returning to reality.

[Overall Style]

Hollywood-style hyperrealistic war sci-fi short film aesthetic—with a real protagonist and real animal, cinematic widescreen format, black borders at the top and bottom, low-saturation dark tones, narrative use of cool and warm colors, slight film grain, shallow depth of field, and crisp facial details.

Tears, scars, animal fur, feathers, scales, or other body surface details, bite marks, tactical clothing, and environmental dust all have a highly realistic texture.

In the real-world setting, the animal already has zombie-like grayish-white, cloudy eyes from the first 1–3 seconds of appearance; the 6–8 second segment is solely a close-up of the animal’s face and zombie-like eyes, without showing the eye mutation process.

The 8–10.2 second memory scene represents a normal, warm, everyday moment before infection occurs—the animal remains healthy, normal, affectionate, and in its natural state.

After leaving the memory, the scene must clearly cut back to the real-world ruin. Only after returning to reality does the zombie-like animal attack the protagonist’s neck, and the neck bite wound triggers infection and mutation.

[Time and Plot Logic Requirements]

The entire plot sequence must strictly adhere to the following order:

Warm daily memory

→ Memory ends completely

→ Cut back to real-world ruin

→ Zombie-like animal reappears

→ Animal attacks the protagonist’s neck

→ Protagonist’s neck develops a bite wound

→ Bite wound initiates infection

→ Black veins spread

→ Protagonist’s eyes turn milky-white

→ Protagonist completes infection and mutation.

During the memory, only warm, everyday interactions between the protagonist and the animal are allowed—no aggression, biting, chasing, panic, injury, bleeding, or mutation is permitted.

The animal must not attack the protagonist during the memory scene. It’s forbidden for the animal to transition directly from hugging, cuddling, licking, or rubbing to biting the protagonist’s neck.

The warm memory must first end completely, then the scene must cut back to reality via a hard cut or motion-matching transition. After returning to reality, the animal reverts to its zombie-like state and suddenly attacks the protagonist’s neck.

The protagonist’s infection and mutation must result from the animal’s attack on the protagonist’s neck in the real-world setting. The mutation must begin precisely from the bite wound site—no mutation can occur out of thin air, no mutation can appear before the animal’s attack, and no bite wound can appear first and then be supplemented by the animal’s attack.

Grief

Generate a hyperrealistic 12-second video with 4K high-definition, cinematic-quality texture. {Camera shooting pixel parameters: 1.6}, natural image noise level: 3.0, {Realism parameter: 1.6}. The light and shadow transitions are seamless, and the human body proportions are perfectly accurate. The protagonist’s facial features, skin tone, and hairstyle are 100% based on Reference Image 1; the protagonist’s facial features, skin tone, and hairstyle are also 100% based on Reference Image 2.

[First Frame]

A fixed wide-angle shot of a somber, cold-toned seaside cliff scene: Under a heavily overcast sky filled with thick gray clouds, the deep gray-blue sea surface ripples with tiny whitecaps, while a rugged, dark-black volcanic rock cliff platform extends toward the sea. On the left side of the cliff edge stands the female protagonist, wearing a long white lace dress with a trailing skirt. Her hair is blown back dramatically by the strong sea breeze, and the extra-long trailing skirt is lifted by the wind, floating outward toward the cliff’s edge. She’s turning her head slightly to face the figure on the right. On the right side of the platform stands the male protagonist, dressed in a long black trench coat paired with a light-colored collared shirt underneath, dark-colored trousers, and black ankle boots. His hair is similarly tousled and flying in the wind. He’s facing the female protagonist, locking eyes with her. The sea breeze is intense, creating an overall atmosphere of oppressive sadness and emotional tension, brimming with farewell sentiment.

[Video—12 Seconds]

The total duration is 12 seconds—a short, emotionally charged cinematic narrative video. The lighting is cold, cloudy, and starkly natural, set against a coastal volcanic cliff backdrop. Throughout the entire video, the characters’ positions remain strictly as follows: on the left cliff edge stands the female protagonist, wearing a long white lace dress with a trailing skirt; on the right platform stands the male protagonist, clad in a long black trench coat, a light-colored collared shirt underneath, dark-colored trousers, and black boots. There are no gender-specific terms used anywhere in the video. The overall atmosphere is heavy and melancholic, full of emotional tension reflecting self-farewell:

0–1s: The camera smoothly zooms from a wide shot to a medium shot. The sky is covered with thick gray clouds, the deep gray-blue sea surface churns with whitecaps, and the rugged, dark-black rocky cliff platform stretches out before us. The two characters stand facing each other, locked in a gaze. A powerful sea breeze blows, lifting the female protagonist’s white trailing skirt and both their hair, which continues to fly backward. The female protagonist, wearing the long dress, has slightly parted lips as if she’s about to speak. Both characters wear solemn expressions, their eyes filled with reluctance and inner turmoil. The cold-toned lighting remains steady throughout.

Cut to close-up

1–3s: Close-up shot of a single character’s face, fully focused and sharply blurred in the background, rendered in a cold gray tone. The female protagonist, wearing the long dress, has tears welling up in her eyes. One tear slowly rolls down her cheek, her lips slightly parted, conveying a sense of sorrow and vulnerability.

Cut to close-up: Another close-up of the male protagonist’s face, fully focused and blurred in the background, rendered in a cold gray tone. His expression is firm yet restrained, his lips tightly pressed together. The focus is locked on the details of the tear tracks on his face.

3–6s: Close-up shot of a single character’s face, with hair blowing softly across the face. Tears continue to roll down her cheeks. The female protagonist’s expression gradually shifts from sadness to a sense of release as she looks directly into the camera.

Cut to close-up: Another close-up of the male protagonist’s face, with hair blowing softly across the face. His expression shifts from restraint to determination, as his emotions move from conflict to resolve.

6–7s: Close-up shot of hands, with the background blurred into a hazy gray-white tone. The female protagonist’s hand, adorned with lace-trimmed cuffs, and the male protagonist’s hand, revealing the cuff of his shirt, are about to touch. Then, the male protagonist’s hand slowly pulls back, and the two hands ultimately miss each other, clearly showing the texture of the cuffs and the movement of the hands.

7–9s: Medium-shot, fixed camera angle. The male protagonist raises his right arm, palm facing forward, making a pushing motion. The female protagonist, pushed by this force, leans violently backward, losing her balance and falling off the cliff edge toward the sea. Her white dress billows open completely in the air, spreading out gracefully before quickly falling out of frame and disappearing. The male protagonist remains standing still, his palm still extended, watching the direction where the female protagonist fell. His expression is calm and resolute, while the female protagonist’s expression at the moment of her fall is one of serene acceptance, without any sign of panic.

9–12s: Wide-shot, fixed camera angle. After the female protagonist completely disappears, the male protagonist slowly retracts his outstretched hand. He stands silently for two seconds, then turns his back to the camera and walks away along the cliff platform toward the right side of the frame, eventually stepping out of the frame entirely. The final shot leaves only the dark cliff and churning sea under the gloomy sky, creating an empty and desolate atmosphere. Throughout the entire sequence, the sea breeze feels genuinely dynamic, the skirt and hair flutter naturally, delivering a cinematic-quality texture and ensuring a smooth, natural progression of emotions.

Falling Vows

Replace the pet in the video with the one shown in Figure 1, preserving all full-body features and the tail; keep both the scene and the overall styling unchanged from the original video.

Pet Shot Motion

Generate an 8-second hyper-realistic video with 4K, cinematic-quality texture. Camera pixel settings: 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 product itself is 100% based on the details, logo, and shape of the provided image.
[8-Second Video]
Camera Angles and Scene (100% preserving the original live-streaming environment):
The entire shot uses a fixed, eye-level camera position, subtly capturing the very slight, natural breathing‑induced shake unique to live streaming—no pushing, pulling, panning, or tilting. The framing is medium-close up, with a black ring-shaped fill light and a white live‑streaming smartphone mounted centrally on its stand always visible in the foreground, creating a natural framing effect. Behind the model stands a lit makeup mirror adorned with warm white circular LED beads, against a pale beige wall, with small yellow flower accents in the corner. An European‑American female model faces the phone, standing in an empty space to showcase the clothing while wearing the full outfit. Lighting consists of soft, front‑facing fill light that matches the live‑streaming room’s illumination exactly.
Action Sequence by Second (with the garment fully worn):
0–2 seconds: Opening interaction
The model wears the featured product, standing facing the camera with a gentle smile directed at the phone lens, waving hello to viewers in the live stream with a relaxed, friendly demeanor.
2–4 seconds: Silhouette demonstration
The model turns slightly to her left, gently pinching fabric at her side waist and lightly tugging it outward to clearly highlight the silhouette and the drape of the material, keeping her head oriented toward the camera throughout.
4–6 seconds: Detail showcase
Returning to face the camera, the model lifts her left hand to delicately tug at the sleeve cuff, revealing the hemming and tailoring details. She leans ever so slightly toward the camera, allowing the fabric’s nuances to come into sharper focus, all while maintaining a lively, natural expression.
6–8 seconds: Closing presentation
With both hands resting naturally along the hemline, the model gently smooths the bottom of the garment to emphasize its length, then tilts her head toward the camera and flashes a bright, cheerful smile, concluding this segment of the on‑body presentation. The movements are unhurried and natural, perfectly mirroring the authentic rhythm of everyday apparel livestreams.

Seedance 2.5 AI Video Templates

Explore AI video templates powered by Seedance 2.5.