βͺ AI is making reality look fake. ποΈ
βͺ AI video β’ Seedance 2.5 β’ Free prompts & workflows
βͺ I create the unreal. You decide whatβs real.
βͺ Collabs? DM me π€
@YourAlphaMom@makeugc The real test isnβt the best result after 10 attempts. Itβs the best result you can get consistently on the first try. Thatβs where the economics of AI video get interesting π
She was the calmest robot at the expo. For about 4 seconds.
Made with Unlimited Seedance on @makeugc
Promptπ
Create a native 15-second, vertical 9:16 handheld iPhone video filmed by a person standing in a crowd at a tech expo booth. One continuous shot, no cuts. The uploaded image is the exact first frame of the video.
REFERENCE IMAGE LOCK
The android girl in the center of the image is NOVA. Keep her 100 percent identical to the image in every frame: the same face, facial features and proportions, the same light hazel eyes, brows, lips and skin tone, the same jet black hair with center part, the same teal crop T-shirt, the same high-waisted black bike shorts with silver buttons, the same body shape, and the same exposed mechanical waist band with silver pistons, springs and servos. Do not redesign, beautify, age or slim her. Keep the same booth, crowd, ceiling, light, colors and camera angle as in the image.
The spectator hand holding sunglasses on the LEFT edge lowers out of frame within 0.5 seconds and never returns. Nova wears no sunglasses in the entire video.
CORE RULES
One continuous handheld shot. One main character: NOVA. Nova stays on the same spot for the whole video: she never walks, never turns around, never spins. From 00:04.0 there is an empty circle of at least 3 meters around her; nobody is within 3 meters of her while she moves. She never touches anyone. Exactly two people fall, and they fall from fright at the loud spark crack, while Nova is still standing frozen and not moving. Nobody falls while Nova is striking. Everyone else stays standing for the whole video. Nobody is injured. No text, captions or overlays on screen. All people are adults aged 18 to 30.
CHARACTER
NOVA looks exactly like the image: a female android who looks like a real young woman, apparently 23 years old, slim, very narrow waist, full, large, natural-looking bust, long athletic legs, realistic body proportions. Straight glossy jet black hair, center part. Fitted teal short-sleeve crop T-shirt, high-waisted black bike shorts with a row of small silver buttons, black sneakers.
THE ROBOT WAIST: exactly as in the image, a fully exposed band between top and shorts with brushed silver steel actuators, vertical pistons with coil springs, small servo motors, black cables and aluminium plates. Always visible, never changes shape. In fight mode tiny blue status lights blink inside it.
FACE
Identical to the image in every frame. Normal mode: sly closed-lip half smile, eyes open and on the camera; she never looks down. Fight mode: blank, emotionless machine face, eyes locked on the camera, lips closed. Only upper teeth if teeth show. Her face is always sharp and never motion blurred. Her hair never covers her face.
VOICE
Soft, calm, slightly synthetic female voice with a small digital edge. One line only, at the end.
CAMERA
Held by a spectator about 2 meters in front of Nova at chest height, pointing slightly down, as in the image. At 00:04.0 the camera person steps back about 1.5 meters once, so Nova is visible from head to feet inside the empty circle with the crowd around, and stays there to the end. A young woman in a denim jacket holding a black tablet is on the RIGHT edge at the start.
CAMERA HAND
The phone holder is never visible.
HANDHELD
Natural small shake in the calm part, a strong jolt at the spark crack, then a slightly nervous shake that keeps Nova centered. No zoom.
LOOK
Plain iPhone video at 60 fps, default phone processing, light sharpening, natural noise, no filters, no color grading. Real documentary feel.
LIGHTING
Even overhead light from rows of round LED ceiling lamps in a large exhibition hall. Neutral cool white, soft shadows, her face always well lit.
VISUAL MATCH
The image must look exactly like the uploaded image: a phone video from inside a dense crowd at a tech expo, neutral grey and white hall, open steel truss ceiling with many round lamps, the teal top as the only strong color accent, black T-shirts with blue lanyards on the crowd, blue panels behind, everyday iPhone sharpness, no overlays.
SETTING
The same robotics booth inside a huge modern exhibition hall during the day.
BACKGROUND
As in the image: about 20 young adults with blue lanyards around Nova, several filming with phones, a metal cart with robot parts behind her, white booth walls, a dark blue backlit poster on the LEFT, a blue flag, steel truss ceiling. T-shirts are plain black with no logos or text. One distant movement: a person walking far behind the booth.
MOTION
Normal mode: human-like but slightly too precise. Fight mode is a malfunctioning robot running a martial arts program: every move is a sharp mechanical jerk that snaps into a pose in 0.15 seconds and then stops dead and holds completely still for 0.4 seconds before the next move, like a robot hitting fixed positions. No flowing dance moves, no swaying, no circling arms. Each snap has a servo click and a whoosh. Hair swings on each snap and settles during each hold.
DEFECT PREVENTION
Hands: Nova's hands are either tight fists with the thumb outside or flat open palms with fingers together, five fingers each, never merging. Every strike goes into empty air in front of her or to her side, never toward a person and never toward the lens. The woman on the RIGHT holds the tablet with both hands until she steps out of frame at 00:03.8. The camera hand is never visible.
Face: identical and sharp in every frame. Her head never jerks or snaps; she always faces the camera, head turns under 20 degrees. Hair never covers her face.
Body and legs: exactly two arms and two legs. Her feet stay on the same spot all video. Every kick uses one leg while the other foot stays flat on the floor carrying her weight; knee up first, then a straight extension forward no higher than waist height, then the foot back down to the same place. No jumps, no spins, no turning her back, no walking. Knees bend only forward over the toes.
People: from 00:04.0 nobody stands within 3 meters of Nova. Exactly two young men end up sitting on the floor, both at the far edge of the circle, at least 3 meters away from her. All other people stay standing for the whole video. Nobody appears from behind Nova.
Objects: one tablet, one metal cart (always behind Nova at 3 meters, untouched). All phones stay in their owners' hands. Objects never vanish, duplicate or pass through bodies. No readable text anywhere.
Continuity: Nova's face, outfit, hair, mechanical waist, booth, LEFT/RIGHT positions and light stay identical through the whole video. Her final pose matches the first frame.
Physics: one spark moment only, a burst of small blue sparks from her waist at 00:03.5. No smoke, no fire. The background never warps.
STORYBOARD
00:00.0-00:02.0 HOOK, FIRST FRAME: exactly the image. Nova faces the camera, sly closed-lip half smile, eyes on the lens, arms along her thighs, robot waist fully visible, crowd filming, chatter. 00:00.0-00:00.5: the hand with sunglasses on the LEFT lowers out of frame. 00:00.5-00:01.3: her smile widens slightly, one slow blink. 00:01.3-00:02.0 STRANGE DETAIL: head and face stay completely still and sharp, smile unchanged. Only her RIGHT shoulder twitches up twice, fast and mechanical, with two sharp electric clicks, and a piston in her waist jerks.
00:02.0-00:03.5 FREEZE: she is too still, smile frozen, eyes not blinking. 00:02.3: her RIGHT arm rises stiffly straight forward to shoulder height, palm flat down, and freezes, rising servo whine. 00:02.9: her LEFT arm rises the same way. Her smile switches off; face blank. The operator on the RIGHT taps the tablet fast. Chatter drops.
00:03.5-00:04.5 SPARK AND PANIC: a burst of small blue sparks pops from her waist with a loud electric crack, blue lights blink inside the machinery. Nova does not move at all, arms still frozen forward. The crowd panics and pushes back hard, opening a wide empty circle around her. The operator ducks out of frame. The camera person jumps back 1.5 meters with a strong jolt. In the push, two young men at the far edge of the circle, one on the LEFT and one on the RIGHT, both at least 3 meters from Nova, lose their balance and drop down sitting on the floor on their backside, hands behind them. They stay sitting on the floor until the end, staring at her. Nova is still frozen and far from them.
00:04.5-00:06.5 SNAP 1: Nova snaps both fists to her hips and drops into a wide horse stance in 0.15 seconds, then holds still. Snap: RIGHT straight punch forward into empty air, stop dead, hold. Snap: LEFT straight punch forward, stop dead, hold. Each with a servo click and a whoosh.
00:06.5-00:08.5 SNAP 2: snap: RIGHT open-palm chop out to her RIGHT side at shoulder height, hold. Snap: LEFT open-palm chop out to her LEFT side, hold, both arms now out to the sides like a T. Snap: both fists back to her hips, hold.
00:08.5-00:10.5 SNAP 3, KICK: weight onto her LEFT leg, RIGHT knee snaps up to waist height, hold 0.3 seconds, the leg snaps straight forward into a front kick into empty air at waist height, stops dead, hold 0.4 seconds, the foot returns to the floor in the same place. Back in the stance. The crowd gasps from a safe distance; the two men on the floor scoot back a little.
00:10.5-00:12.0 FINAL POSE: snap: RIGHT straight punch forward, LEFT fist at hip, stop dead and hold completely still. A fading servo whine. The blue lights in her waist go off. Silence. Everyone frozen with phones up.
00:12.0-00:15.0 PUNCHLINE: she lowers her fist, straightens up out of the stance, feet together, arms along her thighs, exactly like the first frame. Her face switches back to the sly closed-lip half smile, eyes on the camera. 00:12.5-00:14.0 she says calmly: "Sorry. Self defense mode was ON." Slow and playful, stress on ON; not monotone, not shouting. 00:14.0-00:15.0: one slow blink, the crowd bursts out laughing, one man on the floor raises a thumb up.
STAYS STILL: Nova's feet position, the booth walls, poster, cart, ceiling and lights.
SOUND
Busy expo hall ambience and chatter. Two sharp electric clicks at 00:01.3. Rising servo whine at 00:02.3. Loud electric crack and gasps at 00:03.5, then shouting. A servo click and a short whoosh on every snap. Silence at 00:11.0. Her line, then a burst of laughter. No music.
HANDS AND FEET
Fists tight with the thumb outside, open palms flat with fingers together, five fingers each. Supporting foot always flat on the floor. Sneakers stay on. No close-up of feet.
CONTINUITY AND REALISM
Same Nova as the image in every frame, same outfit, same visible mechanical waist, same booth, same light. One continuous take. The crowd behaves like real expo visitors: scared, keeping their distance, then laughing.
FINAL FEEL
Real phone footage from a robotics expo where a lifelike android girl short-circuits into a rigid martial arts program, the crowd panics, two guys end up on the floor, and she calmly apologizes. Not polished influencer content.
The bear came for the candy. Nobody told him the doll wanted a hug
Made with Unlimited Seedance on @makeugc
Promptπ
SEEDANCE MASTER PROMPT: "THE DOLL THAT SCARED THE BEAR"
Photorealistic Halloween doorbell-camera incident | Target duration: 23 seconds | One continuous security-camera viewpoint
CORE OBJECTIVE
Generate a highly believable, unexpectedly funny-but-creepy Halloween incident that looks like genuine consumer doorbell security footage accidentally capturing something impossible. The clip must feel observed rather than directed: a real suburban porch at night, a real-looking black bear behaving like a curious wild animal, an ordinary creepy Halloween doll, a bowl of candy, a glass front door, and a homeowner visible behind the glass. The doll's sudden movement is the single main shock. The bear's startled escape is the physical payoff. The final image of the doll holding the homeowner's phone is the quiet, uncanny twist. Keep the action clear enough to understand instantly, with no explanatory dialogue, no cinematic coverage, no dramatic music, and no obvious AI visual artifacts.
The clip should resemble authentic doorbell-camera footage viewed after a motion alert. It must not resemble a polished horror-movie trailer, a music video, a game cutscene, or a sequence of unrelated generated shots. Use one fixed camera, one spatially consistent porch, continuous real-time action, practical night lighting, natural animal movement, and restrained visual storytelling. The camera never becomes a character and never moves to help the audience. The viewer is simply watching a security recording.
FORMAT AND VISUAL LANGUAGE
- Duration: approximately 22-25 seconds.
- Aspect ratio: vertical 9:16 composition, as if exported from a vertical smart doorbell camera. Preserve the full porch action inside this frame.
- Camera: a single fixed, wide-angle doorbell security camera mounted beside the front door, about 1.6-1.8 metres above the porch floor, angled slightly downward.
- Lens: realistic consumer security-camera wide angle, approximately 150-160-degree field of view. Mild edge distortion is acceptable, but do not bend bodies, railings, or door frames unnaturally.
- Recording character: compressed low-light digital footage, subtle sensor noise, slight sharpening, modest compression artifacts, realistic shadow noise, and believable automatic exposure. The image should still be clear enough to read the doll's eyes and the bear's reaction.
- Lighting: a warm porch bulb above the door, a weak spill of indoor light through the glass, and cool ambient darkness beyond the steps. The porch bulb is the main light source. Do not add theatrical spotlights, coloured horror lighting, artificial fog, or glowing supernatural effects.
- Colour: natural night-time colours, muted wood and siding, warm amber near the door, cooler shadows in the yard. Avoid extreme teal-orange grading, crushed blacks, oversaturated orange, or a glossy commercial finish.
- Audio: raw security-camera microphone sound. Quiet night ambience, faint leaves moving, subtle bear snuffling, candy-wrapper rustle, small plastic-bucket sounds, claws or paws contacting the porch, one abrupt startled animal vocalization if physically plausible, quick scrambling footsteps, and the faint sound of the homeowner reacting behind the glass. No soundtrack, no narrator, no captions, no voice-over, no added cinematic bass hit, and no artificial sting.
- The timestamp or security-camera interface may be minimally visible, but it must not distract. Do not generate legible brand logos, fake app notifications, random text, watermarks, or prominent UI overlays.
LOCATION AND BLOCKING
The location is an ordinary American suburban front porch decorated for Halloween. The door and its frame occupy the right side of the vertical image. The left and centre contain the porch floor, a small candy bucket, a seated life-size Halloween doll, and the top of two or three steps leading down into a dark front yard. The house siding and porch railing should remain fixed and geometrically consistent throughout the clip. Through the glass panel or glass in the front door, the homeowner can be seen inside the house, partially obscured by reflections and indoor darkness.
The porch should look lived-in and ordinary, not like a haunted attraction. Include a few restrained Halloween details: a small orange candy bucket with a simple pumpkin face, one or two small pumpkins, a short string of dim warm lights if naturally placed, and a life-size doll in a pale vintage dress. Do not clutter the scene. The candy bucket is placed on the porch floor near the doll, close enough for the bear to investigate without crossing the whole porch. A smartphone belonging to the homeowner rests near the threshold, on a narrow ledge or on the floor just inside the door's exterior edge, visible enough that the doll can plausibly reach it after the bear leaves. Keep the phone's position fixed until the final beat.
The doll is unsettling but physically plausible as a Halloween prop: a life-size child-sized figure with a worn pale face, old-fashioned light dress, dark hair, glassy eyes, and a slightly open neutral mouth. It should look like a real animatronic or antique-style Halloween decoration, not a supernatural creature. The hands are articulated prop hands, capable of a quick mechanical movement. The doll sits on a low porch chair or sturdy stool beside the candy bucket. Its feet, skirt, chair and hands must stay spatially consistent. At first, the doll is completely still. No blinking, no head turns, no torso motion, no moving hair, and no hint of movement before the eyes begin tracking the bear.
The bear is a realistic adult black bear of believable size and anatomy. Its fur is dark and textured, with natural highlights from the porch bulb. It moves cautiously, sniffing and investigating food, with appropriate weight transfer and four-legged gait. Its paws must contact the porch surface naturally. The bear must not behave like a human, grin, stand theatrically, or perform exaggerated cartoon movements. The startled response should be abrupt but physically credible: it jerks back, loses its balance for a fraction of a second, knocks the lightweight bucket over, then scrambles down the steps and disappears into the dark yard. Do not show injury, blood, or prolonged aggression. The doll's grab is a brief surprise action, not a sustained attack.
The homeowner is an adult visible behind the glass front door. They wear ordinary casual indoor clothing and remain inside. Their face should be only partly visible through reflections, but their reaction is readable: they witness the sudden grab, recoil, and instinctively cover their mouth. Do not make them open the door, step outside, speak directly to camera, or block the central action. Their movement should be small and believable, not a staged theatrical scream.
CONTINUITY RULES
The entire clip is one uninterrupted recording from the same fixed camera. No cuts, no angle changes, no zooms, no slow motion, no handheld movement, no re-framing, no time jumps, no transitions, and no flash frames. All characters and props occupy the same positions from start to finish unless moved by a visible physical action. The porch architecture, door frame, chair, steps, bucket, pumpkins, phone, and lighting must not change between frames. The bear's size and coat must remain consistent. The doll's face, dress, hair, hands, and body proportions must remain identical throughout. The homeowner remains behind the same glass door. The smartphone cannot teleport, duplicate, vanish, or change colour. The tipped candy bucket remains tipped after it is knocked over. Any spilled candy stays where it lands.
The camera must not pan toward the bear, follow it down the steps, or automatically zoom. The bear can leave the frame naturally through the lower-left or lower-centre steps. The final beat must still be readable in the same fixed composition: the bear has gone, the doll remains in its chair, and the doll is now holding the phone. Avoid any shot where the phone appears in the doll's hands before the bear has left. The phone must be physically reachable from the doll's seated position. If needed, place it on the low outer threshold or a narrow ledge beside the chair, within arm's reach, rather than several metres away. The doll should not cross the porch or move its chair to get it.
TIMED ACTION: FOLLOW THIS ORDER PRECISELY
0.0-2.5 SECONDS: ORDINARY, QUIET SETUP
Begin already recording, as if the motion alert has just started. The porch is quiet. The bear enters from the dark lower-left edge of the frame or is already just stepping onto the porch at the left edge. It approaches the candy bucket with a natural cautious gait, head lowered and nose working. The doll sits completely motionless beside the bucket. The homeowner is faintly visible behind the glass, standing inside and looking toward the porch. Nothing supernatural happens yet. The first seconds should be mundane and credible, so the later event feels surprising. Do not begin with a jump scare, a sudden close-up, a dramatic growl, or an obvious sign that the doll is alive.
2.5-6.5 SECONDS: THE BEAR INVESTIGATES THE CANDY
The bear reaches the bucket, sniffs it, nudges it gently with its nose, and picks up or mouths one piece of wrapped candy. It may use a paw to steady the bucket, but keep the movement animal-like and simple. The bucket gives a small plastic rattle. The bear's shoulders and back move with natural breathing. The doll remains perfectly still, facing forward. Its eyes do not yet move. The homeowner watches from inside but is not yet alarmed. Maintain the same camera exposure and composition; the warm porch light falls naturally across the bear's muzzle and the doll's pale face.
6.5-8.5 SECONDS: THE EYES TRACK
While the bear's nose is down near the candy bucket, the doll's pupils slowly shift sideways toward the bear. Only the eyes move. The head remains fixed, the chin remains at the same angle, the shoulders and dress remain still, and the hands remain resting in the same position. The eye movement should be subtle but unmistakable when noticed. It should look like a small concealed animatronic mechanism, not glowing supernatural eyes. No glowing pupils, no head rotation, no twitching limbs, no floating, and no exaggerated facial morphing. The movement lasts about one second, then the eyes hold their new direction. The bear does not instantly react; it continues sniffing for a brief moment, which makes the eye movement more unnerving.
8.5-10.5 SECONDS: THE BEAR NOTICES
The bear pauses mid-sniff. Its ears lift slightly, its head rises a little, and it looks toward the doll's face. It shifts its weight back by a small amount, then cautiously leans its muzzle closer as if trying to identify the object. The movement should be slow and curious, not aggressive. The doll's head and body remain absolutely fixed. Only its eyes continue to watch the bear. The room behind the glass stays quiet; the homeowner begins to lean forward slightly, trying to see what the bear is doing. The tension comes from the distance closing between a large real animal and a completely still prop.
10.5-12.0 SECONDS: A SHORT, TENSE PAUSE
The bear brings its muzzle close to the doll's face, stopping just short of touching it. Hold the stillness for a fraction of a second. The bear's nose twitches; its breath and soft snuffling are audible. The doll's eyes remain locked on the bear. The audience should have time to expect nothing more than a strange animatronic trick. Do not prolong the pause beyond what feels natural. The sudden movement must arrive before the tension becomes slow or repetitive. The bear's face, the doll's hands, and the bucket should all remain visible in the wide frame.
12.0-13.0 SECONDS: THE GRAB
In one fast, mechanically sharp action, the doll suddenly lifts both articulated hands and briefly grabs the bear's muzzle from the sides. The movement is startling and quick, like a concealed animatronic trigger firing. Both hands make visible contact with the fur around the muzzle, but do not cover the bear's eyes completely and do not deform the bear's skull. The doll's torso stays seated and anchored to the chair; its hips and legs do not slide, and the chair does not travel across the porch. The hands move from their resting position to the bear's muzzle in a single clear action. Do not create extra arms, elongated fingers, rubbery limbs, or impossible joint angles. The doll does not bite, growl, stretch its mouth unnaturally, or transform into a monster. The shock comes from a recognisable doll suddenly moving and grabbing the animal.
13.0-15.0 SECONDS: THE BEAR'S STARTLED ESCAPE
The bear reacts instantly and believably: a sharp backward jerk, a startled huff or yelp, a rapid pivot, and a scramble away from the doll. It bumps the candy bucket, which tips onto its side and spills several wrapped sweets across the porch floor. The bear's paws skid slightly but retain contact with the ground; its body mass and momentum feel real. It hurries down the steps and exits into the dark yard. The action is chaotic for a moment but never becomes an unreadable blur. The doll releases the bear as it pulls away and returns its hands toward its lap. The doll must not cling to the bear, fly off the chair, or pursue it. Do not show the bear attacking the doll or the homeowner. The escape should feel like a startled wild animal fleeing an unexpected mechanical movement.
15.0-17.5 SECONDS: THE HOMEOWNER REACTS
As the bear disappears down the steps, the homeowner behind the glass recoils from the door. They lift one hand to cover their mouth and take one or two involuntary steps backward. Their expression is startled disbelief, not a theatrical horror scream. Their reflection and the porch light may overlap naturally on the glass. They do not open the door or come outside. The audio captures a small muffled gasp or sharp intake of breath from inside, barely audible through the glass. The porch is now still except for a few sweets settling on the floor and the faint night ambience. The doll sits in its original position, once again completely motionless.
this is exactly the kind of video AI should be making. the robot found the mints
Made with Unlimited Seedance on @makeugc
Prompt:
Create a native 30-second, vertical 9:16 amateur phone video filmed by a passenger inside a stopped car on a highway shoulder in heavy rain. No reference image. Raw, shaky, real-life viral footage that a random driver posted. Everything happens in one continuous, logical, real-time order.
CORE RULES
Vertical 9:16 only. One location: a rainy highway shoulder. One character: a Tesla Optimus humanoid robot. Nobody else appears outside. The camera is a phone held by an unseen adult passenger in the stopped car, looking out of the passenger side window from ONE fixed position for the whole video: same angle, same distance, same side. The only changes are a phone zoom and one tilt up to the sky and back down. The car never moves. Nine shots with hard cuts, in chronological order. No flash-forward. No music. No added captions.
CHARACTER
A Tesla Optimus humanoid robot, adult human size, about 173 cm tall. A sleek glossy white body shell over black joints. The head is a smooth, rounded, glossy black faceplate with no eyes, no mouth and no facial features. The torso front and the hands are black. The word "TESLA" is in clean, widely spaced, thin black capital letters across the white upper chest. Exactly two arms, two legs, and two black hands with five slim fingers each. It moves fluidly and naturally like a relaxed human worker: arms swing when walking, the body bobs slightly with each step, weight shifts from leg to leg, the head turns calmly. It is never frozen and never stands stiff.
FACE
The black faceplate stays smooth and identical in every shot. It never cracks, deforms or shows a face. The "TESLA" letters stay in the same place, the same size and spelled correctly in every shot.
VOICE
The robot never speaks. Off-screen passengers in the car: adult men, American, stunned and laughing, natural and not acted, never shouted. Lines: shot 3, "He's DRINKING it."; shot 5, "No no no... NOT the mints."; shot 7, laughing, "He's GONE."; shot 8, "Where'd he GO?"; shot 9, laughing hard, "He's getting ANOTHER box." Each line is short and spoken slowly.
CAMERA
Vertical phone at the eye level of a seated passenger, looking out of the side window at a slight downward angle. A wide view shows both trucks; the lower part shows the flooded wet road. A soft-focus strip of the dark car door and window frame sits in the bottom-left corner in every shot except shot 7. Shots 2 to 5 use a shaky 2x phone zoom from the same position.
CAMERA HAND
The phone is held by an unseen passenger. Their hands, arms and phone are never visible and never reflected in the glass.
HANDHELD
Constant small handheld shake, slightly late reframes, autofocus sometimes hunting between raindrops on the glass and the scene.
LOOK
Real smartphone footage on a dark rainy day: flat grey, slightly soft, mild compression, light noise, raindrops and streaks on the glass.
VISUAL / LIGHTING
Overcast storm daylight only: a bright white-grey sky, no sun, no hard shadows. Muted greys, wet dark asphalt and a dark green tree line, with two colour accents: the red tanker and the white-and-blue candy truck.
VISUAL MATCH
The image must look exactly like this: a casual phone clip from a stopped car on a rainy highway. A red tanker truck parked on the shoulder sprays a wide, thick fan of brown fizzy cola sideways out of a torn hole in the side of its tank. Behind it, a box truck with an open rear shows stacked candy boxes. The road is flooded with shallow brown water. Handheld shake, soft focus, raindrops on the glass, muted overcast colours. No colour grading, no slow motion, no drone, no gimbal, no cinematic close-ups.
SETTING
A highway shoulder in heavy rain. On the LEFT, a red tanker truck is seen from the side, with large white script letters "FIZZCO" on the tank. Low on the side of the round tank there is a ragged, torn puncture hole about the size of a fist, with jagged metal edges bent inward, as if it had been ripped open in a crash. There is NO valve, NO tap, NO pipe and NO nozzle: the cola bursts straight out of the hole in the tank wall as a wide, thick, foaming fan of brown liquid shooting diagonally down onto the road. On the RIGHT, slightly behind the tanker, is a white box truck with blue candy pictures and the large word "MINTO" on its side. Its rear roll-up door is fully open, showing stacked blue-and-white boxes of round white mint candies.
BACKGROUND
Foreground: raindrops on the glass and the door frame strip. Midground: the trucks, the cola fan, and the shallow flooded road with brown fizzy water. Background: grey highway lanes, a guardrail, a dark tree line, and distant cars with headlights on. In every shot the anchors stay identical: tanker LEFT, candy truck RIGHT and behind, the torn hole in the tanker side.
MOTION
The cola sprays constantly from the same hole for the whole video, weaker after the explosion. Rain falls steadily. The robot always does something: walks, fills, drinks, carries, dumps, flies, lands, walks again.
DEFECT PREVENTION
Hands: shots 2β4: the robot holds one white plastic jug, already open with no cap. RIGHT hand on the handle; in shots 2 and 3 the LEFT hand also holds the jug's bottom. Shot 5: both hands hold one blue-and-white candy box at its sides, palms under it. Shots 6β9: both hands are empty. Five fingers per hand, never merging with objects or the body. The camera hand is never visible.
Face: the black faceplate is always intact. Head turns stay under 45 degrees.
Body and legs: the robot walks with natural steps through ankle-deep water, each foot planted before the other lifts, arms swinging. When landing in shot 9 it lands feet first and bends its knees once. Exactly two arms and two legs. Joints bend only naturally. Feet never slide. The body never stretches or changes shape, including in the air.
Objects: one tanker, one candy truck, one white jug (after shot 4 it lies floating in the water), one empty candy box (after shot 5 it floats in the foam). Candies fall only out of the box. Objects never vanish, duplicate, teleport or change colour. The only readable text: "TESLA", "FIZZCO", "MINTO". The tanker never has a valve, tap or pipe; only the torn hole.
Continuity across cuts: the same robot, trucks, sides, hole, camera position, rain and light in every shot.
Physics and environment: the cola is a wet, brown, fizzy liquid. When the candies hit it, one single giant column of white wet foam erupts straight up from that spot, like a huge fountain, then rains back down as white drops and leaves a wide lake of white foam on the road. Never sand, never dust, never smoke, never fire. No reflections in the glass. Nobody else appears. The background never warps.
STORYBOARD
SHOT 1 β 00:00β00:05.0 (wide)
START POSE: The cola fan is already bursting hard out of the torn hole in the tanker side. The robot walks in from the RIGHT along the flooded road toward the spray, with the white jug hanging from its RIGHT hand, its LEFT arm swinging and its LEFT foot forward.
00:00β00:02.5: the cola sprays loudly and the robot walks with splashing steps.
00:02.5β00:05.0: it reaches the edge of the spray and stops, feet apart.
STAYS STILL: the trucks.
SHOT 2 β 00:05.0β00:09.0 (2x zoom, same angle)
START POSE: The robot stands beside the spray with the jug held at waist height in both hands, opening up.
00:05.0β00:07.0: it pushes the jug into the side of the cola spray and the jug fills; cola splashes over its hands and chest.
00:07.0β00:09.0: it pulls the full jug back out of the spray with both hands.
STAYS STILL: the tanker.
SHOT 3 β 00:09.0β00:11.0
START POSE: The robot stands next to the spray with the full jug at chest height.
00:09.0β00:10.0: it lifts the jug to its faceplate and tilts it back like drinking.
00:10.0β00:11.0: cola pours over the faceplate and runs down the white chest. Voice: "He's DRINKING it."
STAYS STILL: the robot's feet.
SHOT 4 β 00:11.0β00:13.0
START POSE: The robot lowers the jug to its RIGHT side and turns toward the candy truck.
00:11.0β00:13.0: it walks toward the open rear of the candy truck and drops the jug into the water.
STAYS STILL: the trucks.
SHOT 5 β 00:13.0β00:17.0
START POSE: The robot is at the open rear of the candy truck, both hands pulling one blue-and-white box off the stack.
00:13.0β00:14.5: it carries the box against its chest and walks back toward the spray.
00:14.5β00:16.0: it tips the box forward and hundreds of white mint candies pour into the brown water under the spray. Voice: "No no no... NOT the mints."
00:16.0β00:17.0: it drops the empty box and the water starts bubbling white around its legs.
STAYS STILL: the tanker.
SHOT 6 β 00:17.0β00:18.0 (wide)
START POSE: The robot stands in bubbling water, hands empty.
00:17.0β00:18.0: a giant white foam column erupts straight up from under the robot with a roar.
STAYS STILL: the trucks.
SHOT 7 β 00:18.0β00:22.0 (camera tilts up)
START POSE: The robot is on top of the white foam column, already rising, arms and legs flailing.
00:18.0β00:20.0: the camera tilts up and the robot shoots high into the grey sky on the tall white column.
00:20.0β00:22.0: the robot is a small dark figure high above. Voice, laughing: "He's GONE."
STAYS STILL: the sky.
SHOT 8 β 00:22.0β00:25.5 (camera tilts back down, wide, same angle as shot 1)
START POSE: The same wide view of both trucks. The foam column has collapsed into a wide lake of white foam on the road. The empty box floats in it and the cola still sprays weakly from the hole. No robot.
00:22.0β00:25.5: white foam drops rain down and the camera searches a little. Voice: "Where'd he GO?"
STAYS STILL: the trucks.
SHOT 9 β 00:25.5β00:30.0 (same wide view)
START POSE: The same view; the robot drops into frame from above, feet first, straight down into the white foam lake.
00:25.5β00:26.5: it lands on its feet with a big white splash, bending its knees once.
00:26.5β00:28.0: it straightens up, foam sliding off its white shell, and calmly turns toward the candy truck.
00:28.0β00:30.0: it walks with relaxed steps back toward the open rear of the candy truck for another box. Voice, laughing hard: "He's getting ANOTHER box."
STAYS STILL: the trucks, the camera position.
SOUND
Heavy rain, the hissing cola spray, splashing steps, passengers laughing quietly. Shot 6: a deep roaring whoosh. Shot 9: a heavy wet splash on landing, then loud laughter. No music.
HANDS AND FEET
Five fingers per hand, never extra. Feet planted in the water, never sliding or sinking.
CONTINUITY AND REALISM
Physically real and logical. The robot is heavy, and its splashes match its steps. The light never changes.
FINAL FEEL
A real, chaotic, unbelievable moment a stunned driver caught on their phone in the rain. Not polished influencer content.
Gave my horse his first peppermint and i can't stop rewatching his face
Made with Unlimited Seedance on @makeugc
Prompt π
Create a native 17-second, vertical 9:16 self-recorded phone clip filmed as ONE continuous handheld take with no cuts. No reference image.
CORE RULES
One adult woman, LILA, and one horse, BISCUIT, in one unbroken take. No cuts, no jumps in time, no changes of framing, outfit, light or background. LILA has exactly two arms and two legs, five fingers on each hand, joints bend only naturally. She films with her RIGHT hand; that hand and the phone always stay out of frame. Her eyes stay alive, she blinks naturally, her mouth matches her words. BISCUIT has one head, two ears, two eyes and a normal horse mouth. Real-time motion, no freezes. No text, logos, extra people or extra animals.
STORY IN ONE LINE
LILA feeds her horse his first peppermint ever from her flat palm. The cold mint numbs his face: his chewing freezes, his lower lip slowly sags, his tongue falls out and he starts dumbly wobbling his head like a bobblehead. LILA stands beside him filming, laughing and commenting. She never eats anything; after the feeding her hands stay out of frame; the horse is the star.
BEAT RULE
Every action is clear and readable on a phone screen: one main action at a time, with a clear start and a clear end. LILA only watches, laughs and comments; she never copies him. No two big actions overlap. No vague, half-done or blended movements.
CHARACTER
Tone: a real, chaotic barn moment a girl films with her horse, laughing, no posing.
LILA is a beautiful 25-year-old woman, naturally sexy in an effortless, believable way. Honey-blonde hair in a loose low braid over her RIGHT shoulder; two short wisps at her temples, never across her eyes or lips. Slim, curvy figure with a full, large, natural-looking bust and realistic body proportions. A soft mint-green chunky knit crew-neck sweater with a round neckline at her collarbones, no cleavage, small silver stud earrings.
BISCUIT is a calm, almost black dark-bay horse with a glossy coat, black mane, a small white star on his forehead, soft dark eyes and a dark velvety muzzle with fine whiskers. A brown leather halter with brass buckles, always on.
FACE
LILA: oval face, soft cheekbones, small rounded chin, wide grey-green eyes with dark lashes, light-brown straight brows, small slightly upturned nose, full lips with soft pink balm, straight natural white teeth, fair skin with light freckles and a faint pink flush from the cool air. Her gaze moves before her head; she blinks every two to four seconds; expressions change gradually. When she speaks her mouth opens only moderately; lips press together on m, b and p. Smiles show only her upper teeth. Head turns under 45 degrees. No glassy eyes, no frozen expression, no warped teeth, no changing features.
VOICE
Only LILA speaks: light, warm, slightly raspy, recorded by the phone microphone with a little barn echo. Real laughter breaks into her words. She speaks only the eight short lines written in the storyboard. Soft American accent. Not announcer-like, not a voice-over. The last line ends before the video ends.
CAMERA
Ordinary modern smartphone footage, selfie POV, one continuous take. LILA holds the phone in her RIGHT hand at arm's length, slightly below her eye level, the lens looking a little upward. Framing: LILA's face always in the LOWER LEFT third of the screen, BISCUIT's head large in the RIGHT two thirds beside her LEFT shoulder, with headroom above his ears. The phone NEVER follows the horse away from LILA: when BISCUIT lifts his head high, the phone tilts up only slightly and his muzzle may leave the top of the frame, but LILA's whole face always stays in frame. It never zooms, never jumps and never loses either face.
CAMERA HAND
Her RIGHT arm leaves the frame at the bottom-LEFT corner of the screen; never her RIGHT hand or the phone. The frame never cuts her arm at the elbow or wrist.
HANDHELD
Natural hand shake, slightly crooked framing, small corrections, stronger shake when she laughs, brief motion blur, occasional autofocus hunting between her face and the horse. No gimbal-like stability.
LOOK
Plain phone video: mild compression, slightly soft shadows, a little smartphone edge sharpening, natural warm white balance, no filter. Real skin with pores and freckles, real horse hair with dust, fine knit texture. No beauty filter, no plastic skin, no commercial grading.
LIGHTING
Only two sources: a round copper heat lamp with a warm orange glow high in the upper LEFT corner of the screen, and soft daylight from a wide open doorway in the RIGHT background. About one third of the frame in warm shadow. The lamp is never directly above LILA's face. The light never changes.
VISUAL MATCH
The image must look exactly like this: a casual phone video inside a wooden horse stall, the horse's head very close and large, the camera slightly below and looking up, pale ceiling planks across the top, the copper heat lamp glowing upper LEFT, a black metal grille behind LILA, a bright doorway with hay bales behind BISCUIT, warm honey wood and a near-black coat, mint green as the only cool colour. One unbroken take, slight shake, no overlays.
SETTING
A small private stable stall in the late morning. Pale plank ceiling, warm wooden walls, a black metal grille on the back wall behind LILA, a wide open doorway with daylight and stacked hay bales behind BISCUIT. Nobody else, no other horses.
BACKGROUND
Readable in half a second as a horse stall. Three layers: BISCUIT's halter and muzzle closest; LILA's face and BISCUIT's eyes in the middle; the grille and the doorway behind. Background life: dust particles drift slowly in the doorway light. No reflections of LILA or the phone anywhere.
MOTION
Real weight and natural timing: preparation, action, follow-through; her braid swings when she laughs and settles. Eyes lead, then head, then shoulders. BISCUIT's mint reaction builds step by step: first a freeze, then a slow numb sag of his lower lip, then his tongue falls out, then loose, dopey side-to-side head wobbles like a bobblehead, his slack lip and tongue flopping with each wobble, then the classic horse laugh: head high, upper lip curled up showing his upper teeth. His ears flop out to the sides; his body never moves. Something happens every second: no dead frames, no slow or sped-up motion, no stiff posing.
DEFECT PREVENTION
Hands: only her LEFT hand appears, only at 00.0-02.0, entering from the bottom-RIGHT edge, palm up and flat, fingers together, one peppermint on the centre of the palm; after his lips take it, the hand drops out of frame for good. Five fingers; fingers never merge with each other, the mint or his lips.
Mint: one small round flat peppermint disc, about 2 cm, bright white with bold red pinwheel swirl stripes, unwrapped. Never pink, never a candy cane, never a stick, never a pill.
Horse lip and tongue: everything sags naturally and softly, like a real numb horse mouth. His lower lip hangs low and loose, showing a little of his lower front teeth. Only 3 to 5 cm of pink tongue, about the size of a thumb tip, falls out at the side of his mouth, hangs limp and flaps loosely with each wobble. Each time, the tongue comes out of the side named in the storyboard and goes fully back in. Never a huge cartoon tongue, never covering his muzzle; his mouth never opens wide; no inside of the mouth. During the horse laugh at 10.2-12.0 only his upper lip curls up, showing his upper front teeth and pink gum, exactly like a real horse; his jaw stays closed.
Horse head and neck: moves like a real horse; neck bends naturally; ears, star and halter stay the same; his body never moves. Except during the horse laugh at 10.2-12.0, his head stays beside LILA's LEFT shoulder.
LILA in frame: her whole face is always inside the frame, never cut by the edge, even when the horse moves. Her skin tone never changes. She never eats anything and never makes his face.
Continuity: one take, so outfit, braid, studs, halter, star, lamp, grille, doorway and light never change. No readable text.
Physics and environment: only dust moves in the background. Nobody else appears. The background never warps.
STORYBOARD (17s, one continuous take; the numbness builds in this exact order)
00:00.0-00:02.0 - START POSE, the feeding: LILA's LEFT hand is already in frame, palm up and flat under BISCUIT's muzzle, the round red-and-white peppermint clearly visible in the centre of her palm. BISCUIT's velvety lips wiggle eagerly over her palm, his whiskers tickling it. LILA (whispering, excited): "First peppermint. EVER." At 01.6 his lips scoop the mint off her palm; her empty hand drops out of frame.
00:02.0-00:03.0 - BISCUIT lifts his head level beside her, ears forward, staring straight into the lens, crunching the mint loudly, jaw grinding side to side. LILA, grinning, watches him.
00:03.0-00:04.2 - Freeze: mid-chew he suddenly stops, jaw frozen, eyes widen slightly, ears slowly rotate out to the sides. A dead silent stare into the lens. LILA (whispering, delighted): "Uh oh."
00:04.2-00:05.4 - Numb lip: his lower lip slowly sags down, heavier and heavier, until it hangs low and loose, a little of his lower front teeth showing, the rest of his face completely still. LILA (biting back a laugh): "His face is BUFFERING."
00:05.4-00:06.4 - Tongue: the pink tongue slowly slides out at the LEFT side of his mouth and just hangs there, limp. He rolls his eyes down and looks at his own tongue, a little cross-eyed, confused. Then the tongue gives three fast little wiggles in the air, like he is trying to cool it. LILA snorts with laughter; the frame shakes.
00:06.4-00:08.6 - The wobble: he tips his head back slightly, eyes half-closed, and starts loose, dopey side-to-side head wobbles like a bobblehead, his slack lip and limp tongue flapping loosely from side to side with each wobble, like a flag in the wind. LILA bursts out laughing, eyes squeezed into crescents, the frame shaking: "He's having a MOMENT."
00:08.6-00:10.2 - Still wobbling, tongue still flopping, he turns his head toward the RIGHT side of the screen, showing his cheek and the halter buckle. LILA (crying with laughter, mock-serious): "Sir. Your TONGUE."
00:10.2-00:12.0 - The horse laugh: tongue in, he lifts his head high and swings it up and to the LEFT above LILA, neck arching, and curls his upper lip up and back, showing his big upper teeth, like he is laughing at the ceiling; we see it from below beside the copper lamp. He holds it for one second, then drops the lip. LILA (crying with laughter, pointing with her eyes): "Is he LAUGHING at me?"
00:12.0-00:13.4 - He curves his neck back down, head tilted sideways, eye rolling toward LILA. His tongue pops out on the OTHER side of his mouth this time, the RIGHT side, hangs for a beat, wiggles twice and slowly slides back in. LILA (wheezing): "Wrong SIDE, buddy."
00:13.4-00:15.2 - Recovery: he brings his head back level beside her, mouth closed, lip back in place, ears forward, and stares into the lens perfectly calm, as if nothing happened, exactly like at 02.0. Then he blows a long, loud, wet lip flutter straight at the lens, "pbbbbt", his lips vibrating loosely, no spray.
00:15.2-00:17.0 - He nudges his muzzle down toward LILA's pocket below the frame, ears forward, hopeful. LILA (deadpan, to the lens, still shaking with laughter): "And now he wants ANOTHER one." The line ends at 16.8.
SOUND
Quiet stable ambience: a soft breeze, a distant bird, a faint creak of wood. The soft brush of his lips on her palm at 00.0-01.6, BISCUIT's loud crunch at 01.8-03.0, then sudden silence at 03.0-04.2, a soft wet lip sound as his lip sags, his lips flapping during the wobble at 06.4-10.2, a long snort at 09.6, a low throaty breath during the horse laugh at 10.6, a long loud wet lip flutter "pbbbbt" at 14.4. LILA's real laughter, growing from giggles to wheezing. No music.
HANDS AND FEET
Five fingers on each hand. Hands are clearly visible or clearly out of frame. Exactly two arms. No twisted wrists, no extra limbs, nothing passing through objects or the horse.
CONTINUITY AND REALISM
One unbroken take: one woman, one horse, one lamp, one grille, one doorway, always in the same places. No cuts, no floating objects, no sudden transformations, no captions or watermarks. She is an adult.
FINAL FEEL
A real, silly moment a girl filmed with her horse in one go, the kind of clip you send to a friend. Real, imperfect, self-recorded. Not a commercial. Not a fashion shoot. Not polished influencer content.