このモデルで、
これが作れますAI動画20選

For the target video, at 0.00 seconds into the target video, <Picture 1> (from [Shot 1]) is fully referenced.

integrated_multimodal_description: Live-action, cinematic beauty product commercial, vertical 9:16. Preserve the exact frosted glass bottle, the pale amber liquid inside, the gold cap, the water surface, the reflection and the pale pink-beige background from <Picture 1>. The bottle shape, color and cap never change. No text, no logo, no readable characters appear anywhere.

[Shot 1] The small frosted glass serum bottle stands on the calm water surface exactly as in <Picture 1>. The camera pushes in with small amplitude at slow speed. A single clear drop of water falls from above and hits the water just in front of the bottle, sending out perfect circular ripples that pass under the bottle and make its reflection shimmer.

[Shot 2] At 00:04.000, the camera arcs slowly around the bottle to the right with medium amplitude. Soft light travels across the frosted glass and the amber liquid glows from inside. Tiny bubbles rise slowly inside the liquid.

[Shot 3] At 00:07.500, the camera pulls out slightly and stops. The ripples settle, the water becomes a mirror again, and the bottle stands still in soft, even light. All motion settles by 00:09.500.

overall_soundscape: A single water drop, gentle ripples, a soft airy room tone.

non_diegetic_music: N/A

For the target video, at 0.00 seconds into the target video, <Picture 1> (from [Shot 1]) is fully referenced.

integrated_multimodal_description: Live-action, handheld smartphone-style pet commercial, vertical 9:16, warm afternoon backlight in a grassy park. Preserve the exact red-coated Shiba Inu, its face markings, curled tail, red collar, the grass and the soft backlight from <Picture 1>. The dog is the same individual in every moment. No text, no logo, no readable characters appear anywhere, and no people appear.

[Shot 1] The Shiba Inu looks up at the camera and smiles exactly as in <Picture 1>. It tilts its head to one side, then the other. The camera stays close with a slight natural handheld shake.

[Shot 2] At 00:03.500, the dog turns and runs happily across the grass away from the camera, ears back and tail curled. The camera tracks after it at fast speed, low to the ground, grass blurring in the foreground.

[Shot 3] At 00:07.000, the dog turns around, runs back toward the camera and stops right in front of the lens, panting with its tongue out in the golden backlight. The camera pedestals down to the dog's eye level. All motion settles by 00:09.500.

overall_soundscape: Soft paws on grass, light panting, a gentle breeze, distant birds.

non_diegetic_music: N/A

For the target video, at 0.00 seconds into the target video, <Picture 1> (from [Shot 1]) is fully referenced.

integrated_multimodal_description: Live-action, calm and slow lifestyle room film, vertical 9:16, soft morning light. Preserve the exact studio apartment from <Picture 1>: the white walls, the wooden floor, the window with morning sun, the sofa, the small dining table and the green houseplants, in the same positions. Furniture never moves, appears or disappears. No text, no logo, no readable characters appear anywhere, and no people appear.

[Shot 1] The camera starts exactly as in <Picture 1> and pushes in with small amplitude at slow speed, in one single gentle move toward the window for the whole ten seconds. There is no pan, no truck, no arc and no change of direction. The only other motion is quiet and natural: the sheer curtain sways softly in a light breeze, dust particles drift slowly in the beams of morning sunlight, and the leaves of the houseplant move slightly. The sunlight on the floor brightens very gradually. All motion settles by 00:09.500.

overall_soundscape: Quiet room tone, a light breeze through the window, faint birds outside.

non_diegetic_music: N/A

For the target video, at 0.00 seconds into the target video, <Picture 1> (from [Shot 1]) is fully referenced.

integrated_multimodal_description: Live-action, casual smartphone selfie review video (UGC style), vertical 9:16, natural indoor light. Preserve the exact man, his face, hairstyle and dark T-shirt, the small white handheld fan without any text, and the room from <Picture 1>. He is the same person in every moment. His left arm stretches toward the lens holding the phone and stays at the edge of the frame exactly as in <Picture 1>; only his right hand is ever visible, and it holds the fan in every moment of the video. He has exactly two arms. No text, no logo, no readable characters appear anywhere. No subtitles.

[Shot 1] Exactly as in <Picture 1>, he holds the small white fan up beside his face with his right hand and smiles at the lens. Slight natural handheld shake.

[Shot 2] At 00:03.000, the fan blades start spinning. With the same right hand he brings the fan closer to his face; the airflow blows his fringe back and he closes his eyes for a moment with a relieved smile.

[Shot 3] At 00:07.000, he opens his eyes, lowers the fan a little while still holding it in his right hand, looks into the lens and nods twice with a satisfied grin. All motion settles by 00:09.500.

overall_soundscape: A soft whirr of a small fan, light movement of hair and clothing, quiet room tone. The man does not speak.

non_diegetic_music: N/A

Vertical cinematic aerial drone footage flying forward over terraced rice fields in a quiet Japanese mountain valley at sunrise. The terraces are filled with water that mirrors the pink and golden sky, and thin morning mist drifts between the layers.

Seconds 0 to 4: the drone glides low and slowly forward just above the water-filled terraces. The curved edges of each terrace pass beneath the camera, and the reflections of the sky shimmer gently.

Seconds 4 to 7: the drone rises smoothly and tilts up, revealing more terraces stepping down the slope like a staircase, a small cluster of traditional wooden farmhouses with dark tiled roofs at the bottom of the valley, and a few thin lines of smoke rising from them.

Seconds 7 to 10: the sun appears over the far mountain ridge, a warm beam of light spreads across the valley, the mist glows golden, and the drone keeps moving forward in one continuous shot.

Camera: one continuous, stable forward drone move with a gentle rise and tilt. No cuts.

Audio: soft wind, distant birds, the faint sound of flowing water. No music and no speech.

Photorealistic nature documentary footage, natural color, gentle contrast, clear morning air. No text, no logo, no readable characters, no people close to the camera.

Vertical close-up food commercial footage of a Japanese home-style hamburg steak on a small black cast-iron plate, shot on a dark wooden counter in a cozy restaurant kitchen at night, warm tungsten light from the side.

Seconds 0 to 3: extreme close-up. The thick hand-shaped hamburg steak sizzles loudly on the hot iron plate, its browned crust glistening, clear meat juices bubbling out around the edges, and white steam rising in soft curls toward the camera.

Seconds 3 to 6: glossy dark brown demi-glace sauce is poured slowly from above onto the center of the steak. The sauce spreads over the crust, drips down the sides and hits the hot iron, bursting into sharp sizzling bubbles and a new burst of steam.

Seconds 6 to 10: the camera pulls back and slightly up in one smooth move, revealing the whole plate: the steak with sauce, a small pile of sweet corn, three glazed carrot pieces and a sprig of green parsley on the iron plate, steam still rising.

Camera: slow cinematic push-in for the first three seconds, locked for the pour, then a smooth pull-back. Shallow depth of field, crisp focus on the crust and sauce.

Audio: intense sizzling, bubbling juices, the pour of thick sauce, a soft clink of the iron plate. No music and no speech.

Photorealistic food commercial, rich warm color, deep shadows, appetizing texture and steam. No text, no logo, no readable characters, no people and no hands in the frame.

subject_definitions:
<Subject 1> is the girl from <Picture 1> and <Picture 2>, a late-teen anime-style girl with short bright red hair, sharp amber eyes, a black cropped jacket over a white shirt, a knee-length dark pleated skirt, lace-up brown boots and a small belt pouch on her left hip.
<Subject 2> is the rooftop at dusk from <Picture 2>, a high city building rooftop with a metal railing, a water tank, distant city lights below, an orange-to-violet dusk sky and strong wind.

summary:
[reference generation] The target video is a 15.00-second vertical 2D anime action sequence in five shots. <Subject 1> stands on <Subject 2>, leaps off the edge, falls past the lit building facade, lands on the street far below and looks back up at the rooftop she jumped from.

retention_analysis:
<Subject 1> (appears in [Shot 1], [Shot 2], [Shot 3], [Shot 4], [Shot 5]): fully_preserved - the short red hair, amber eyes, black cropped jacket, white shirt, pleated skirt, lace-up boots and belt pouch stay identical in every shot; only pose, camera distance and angle change.
<Subject 2> (appears in [Shot 1], [Shot 2]): fully_preserved - the rooftop railing, water tank, dusk sky colors and distant city lights stay the same place and lighting; in [Shot 5] the same rooftop is only seen far above her.

detailed_description:
The target video is in a polished 2D anime style with clean line art, cel shading with soft gradients on the sky, dynamic speed lines and small light flares; an original look that does not imitate any specific studio or artist. No text, no logo, no readable characters appear anywhere.
[Shot 1] Full body, low angle. <Subject 1> stands at the edge of <Subject 2>, her red hair and jacket whipping in the strong wind, the dusk city far below. The camera pushes in with medium amplitude at slow speed toward her face as she narrows her eyes and looks ahead. Wind roar.
[Shot 2] At 00:03.000, the shot cuts to a side view on <Subject 2>. She sprints three steps and jumps off the edge of the rooftop into the open air; the camera tracks with her at fast speed, speed lines streaming behind her.
[Shot 3] At 00:06.000, the shot cuts to a view looking down along the building facade. She falls feet first past rows of warmly lit windows, her red hair and jacket whipping upward, the street lights growing larger below.
[Shot 4] At 00:09.000, the shot cuts to a low angle on the street at ground level: wet asphalt reflecting shop lights and the dusk sky. She lands on one knee with one hand on the ground; a ring of dust and wind bursts outward and the camera shakes strongly for a moment.
[Shot 5] At 00:12.000, the shot cuts to a medium shot on the street. She stands up, brushes her jacket, and looks back up at the tall building; the rooftop she jumped from is tiny and far above her against the last light of the sunset. She turns her head to the camera with a confident half smile. The camera pulls out with small amplitude at slow speed. All motion settles by 00:14.500.

overall_soundscape:
Strong wind on the rooftop, fast footsteps on concrete, a long whoosh during the fall, a heavy landing impact on the street, then distant city noise.

non_diegetic_music:
N/A

subject_definitions:
<Subject 1> is the woman in <Picture 1>, a Japanese woman in her late twenties with shoulder-length straight black hair, a natural smile, a slim build and white sneakers.
<Subject 2> is the studio in <Picture 1>, a plain white wall and a light gray floor with bright, even studio light.

summary:
[reference generation] The target video is a 10.00-second vertical fashion lookbook sequence in four shots. <Subject 1> stands in <Subject 2> and changes into a different outfit on each snap turn, like a styling lookbook.

retention_analysis:
<Subject 1> (appears in [Shot 1], [Shot 2], [Shot 3], [Shot 4]): partially_preserved - her face, shoulder-length black hair, body shape, smile and white sneakers stay identical in every shot; only the outfit changes as described in each shot, and her pose changes.
<Subject 2> (appears in [Shot 1], [Shot 2], [Shot 3], [Shot 4]): fully_preserved - the white wall, gray floor and bright even light stay exactly the same.

detailed_description:
The target video is in a clean, bright live-action fashion commercial style, full-body framing, sharp focus, natural skin tones. No text, no logo, no readable characters appear anywhere.
[Shot 1] <Subject 1> stands centered in <Subject 2> in a white T-shirt and blue denim jeans, exactly as in <Picture 1>. She smiles and does a quick half turn to her left. The camera is a static shot.
[Shot 2] At 00:02.500, the shot cuts on her turn: she now wears a beige trench coat over a navy knit and cream wide trousers, and completes the turn facing the camera, putting one hand in the coat pocket.
[Shot 3] At 00:05.000, the shot cuts on another turn: she now wears a sage green long sleeveless dress with a thin brown belt, and sways the skirt gently as she steps forward once.
[Shot 4] At 00:07.500, the shot cuts on a final turn: she now wears a black oversized blazer, a white shirt and black slim trousers, and strikes a confident pose with hands on hips, then smiles. The camera pushes in with small amplitude at slow speed. All motion settles by 00:09.500.

overall_soundscape:
Soft footsteps on the studio floor, the rustle of fabric on each turn.

non_diegetic_music:
N/A