An 8-second vertical 9:16 hyper-realistic CGI product advertisement. Sleek, powerful, sophisticated. No on-screen text, no subtitles, no logos, no brand names. Single full-frame shot only. Never a split screen, never a comparison, never model-name labels. @product is the car in the reference image. Lock it exactly: long-hood muscle-coupe silhouette, split navy-blue and brushed-silver body, dual hood stripes, cyan LED outlines along the body and wheel lips, five-spoke silver wheels, cyan-ringed tires, quad headlamps, HUD windshield. Same shape, colours, lights and wheels in every frame. One car, never duplicated. Setting: infinite minimalist virtual studio, deep black void, polished black obsidian floor that mirrors the scene. No people. Lighting: cool-white (6000K) softbox from above. Silver-white and electric-blue light from the sphere, throwing dynamic reflections on the floor. Transformation Start: a single dense sphere, painted in the exact colours of @product — navy blue and brushed silver, with a thin cyan light line around its equator — sitting on the obsidian floor. End: @product, fully solid, settled on the floor. AUDIO — required: sphere rolling on glass-like stone, then a burst, metallic crackles, panels locking, electric-motor hum. No dialogue, no lyrics. 0-2 seconds Already in the scene: the two-tone sphere rolls across the black mirror floor toward camera, rotating so navy and silver catch the light. About one full second of continuous roll. Macro-to-medium, 85mm. 2-5 seconds The rolling sphere bursts outward in zero-gravity. Liquid-metal tendrils in navy, silver and cyan stretch into the wireframe of @product, traced by glowing electric-blue energy lines. 5-7 seconds Metal flows over the energy frame, solidifying into the exact body panels, paint split, cyan light lines and wheels of @product. 7-8 seconds @product gently settles onto the reflective obsidian floor. Headlights flash on. Hold. Negative: extra cars, extra people, readable text, logos, split-screen, hovering drop that never rolls, studio still of the reference photo as frame 1.
Drag the clip sideways
@avatar is the person the vlog follows - it defines face and identity only. Preserve their facial identity, skin tone, hairline, hair, apparent age, build and gender presentation exactly as shown, in every shot; never invent a different face. The attached image outranks this text wherever they disagree. Wardrobe comes from this text, never from @avatar: a light summer top, high-waisted jeans, white trainers and a small crossbody bag, unbranded, identical in every shot. 30-second ultra-photorealistic handheld vlog, 16:9, following @avatar through a warm summer evening at a LARGE amusement park - wide paved avenues, a towering steel roller coaster running overhead, a giant lit Ferris wheel, a drop tower, spinning rides throwing coloured light, neon arcades, food pavilions and thousands of visitors. Not a small country fair. Natural evening light going from gold to deep blue, realistic crowd behaviour, physically accurate motion, no CGI look. Hard cuts between shots, no dissolves, and the film opens already in motion. @avatar walks in through the main gates as the last golden light fades behind the coaster and says, "Okay, this place is huge." Hard cut to a Ferris wheel gondola climbing, the whole park opening up below. Hard cut to a food pavilion where @avatar breaks off a piece of funnel cake, powdered sugar falling like snow. Hard cut to a lit game booth: @avatar throws a ring, it spins and clatters off the bottles, and they laugh and throw again. Hard cut to the main avenue among families and vendors, stopping at an outdoor stage where a live band plays - "This is my favourite part." Hard cut to fireworks bursting over the Ferris wheel, the face of @avatar lit by the flashes as smoke drifts across the colours. Natural amusement-park sound only: the roar of the coaster, ride motors, crowd chatter, a distant PA, the band, and fireworks at the end. The two short lines are spoken in American English and synchronised to the lips. No narration. No subtitles.
A warm, unhurried documentary-style clip following a woman who runs a tiny neighborhood pasta shop. Step by step, in clean hard cuts, she kneads the dough, rolls it flat, cranks it through an old pasta machine into fresh tagliatelle, boils it, tosses it with tomato sauce, and plates it for her one regular customer. Shot with the soft, slightly imperfect look of an early-2000s home camcorder — natural light, ambient street sounds, no music — ending on a quiet "Buon appetito" as the recording cuts to black.
A woman serves to open a beach volleyball rally against her opponent, and the camera locks onto the ball for the entire exchange. After her serve and toss, the shot becomes one continuous take: the camera whips rapidly left and right between the two players as they trade a dig, a set, a mistimed spike and a final hard-driven kill shot into the sand — fast, breathless, and shot like it was caught live courtside rather than staged.
ONE HANDHELD SELFIE-STYLE UGC VIDEO. VERTICAL.ONE PERSON ONLY. ONE CONTINUOUS TAKE. ASSET MAP - every reference is bound here and cited by tag below: @avatar - the person. This is who is on camera; face, hair, skin and body come entirely from @avatar and never from this text. @image_1 - the reference sheet. It shows the same one device twice: closed on the left, and opened out flat on the right. The device in this film is that device, in those two states. CHARACTER LOCK - HIGHEST PRIORITY: the person in every frame is exactly the person in @avatar. Preserve their facial identity, facial structure, eye shape and spacing, eyebrows, nose, mouth, exact skin tone, hairline, hairstyle, hair length, texture and colour, freckles and distinctive marks, apparent age, build and body proportions, identically from the first frame to the last. Preserve their gender presentation exactly as shown; do not masculinize, feminize or otherwise reinterpret it. Do not beautify, slim, de-age or restyle them. No face changes, no identity drift, and no other people anywhere in frame. ATTACHED IMAGES ARE MANDATORY AND FINAL: the attached references outrank every word written below them. Wherever a reference and this text disagree, follow the reference. DEVICE LOCK: the object in their hands is exactly the device in @image_1 - closed as it is shown on the left of that sheet, open as it is shown on the right. Its shape, proportions, thickness, materials, finish, colours and every printed and engraved detail come entirely from those two references and never from this text. Closed, it is visibly thicker and wider than an ordinary single-screen phone, exactly as thick and as wide as it is on the left of @image_1. Its two halves are stacked face to face like a shut book standing on its end, and the hinge spine runs vertically down one long side. Open, it is one single wide screen that fills both hands, with a faint crease running down the middle of the glass, exactly as on the right of @image_1. Never redesign it, never re-letter it, never recolour it, never add a logo, badge, label, sticker, lettering or symbol that is not already on the references, and never remove one that is. There is only one device in the whole video. THE TWO OUTER FACES ARE NOT THE SAME - this matters: when the device is closed it has two different outer faces. One outer face is the back: the raised camera module and the engraved logo, exactly as on the left of @image_1. The opposite outer face is the cover display: a lit glass screen showing a wallpaper, with no camera module and no logo on it at all. The camera module exists on one face only. Never repeat or mirror the camera module onto both faces, never put the logo on both faces, and never let the closed device look identical from both sides. When it is turned over in the hand, the camera side and the screen side must read as clearly different. HOW IT OPENS - READ THIS CAREFULLY: the hinge is a single vertical spine running down one long side, and the device opens sideways like a book. Closed, the spine is on one side and the free opening edge is on the other. To open it, a thumb goes on the front of the free half and the fingers go behind it, and that half swings outward to the side, away from the spine, until both halves are flat and side by side. The finished open device is one wide screen with the crease standing vertically down the middle. It never opens from the bottom edge, never opens from the top edge, never unfolds top over bottom, and is never handled as a clamshell flip phone. No half ever slides, peels, hinges horizontally or swings downward. The spine stays on the same side for the whole clip. HANDS: both hands stay in frame while the device is handled. Exactly five fingers on each hand, correct joints, correct thumb placement, natural grip pressure. No extra fingers, no fused fingers, no missing fingers, no bent or rubbery hands, and the device never floats, clips through a hand or changes size at any point. WARDROBE (from this text, never from @avatar): plain everyday indoor clothing with no logos, no lettering and no visible branding. Unisex cut. The reference shows their face and build, not this outfit - fit it to their body naturally without changing its cut, proportions or colour. LOOK: shot on a phone front camera at arm's length, held by the person themselves. Slightly unsteady handheld framing that drifts and corrects, natural indoor daylight from a window off to one side, an ordinary lived-in room softly out of focus behind them, plain phone-camera colour with no grading and no cinematic lighting. Real skin texture, real pores, no beauty filter, no ring light, no studio look. It must read as a clip somebody actually filmed for social, not as an advertisement. The person is pleased and a little smug the whole way through, like someone showing off something they got first. 0-3s: the person holds the device closed in one hand, up near their face, and starts talking straight to the lens. They turn it slowly so the camera catches its closed edge and its thickness. 3-6s: they bring the second hand in, take hold of the free long side, and swing the two halves apart sideways around the vertical spine, opening it like a book until it lies flat. Still talking, eyes flicking down to it and back to the lens. 6-13s: they hold it fully open in both hands beside their face, screen toward the lens, and slowly tilt and rotate it in the daylight. The screen is switched on the whole time and shows its own wallpaper brightly and clearly. As it tilts, the glass stays matte and readable: the room is never mirrored in it, no window, trees, ceiling or reflection of the person ever appears on the screen, and it never turns into a mirror or blows out into a white glare. They keep talking to the lens. 13-17s: they bring the open screen closer to the lens so it fills more of the frame, turning it a few degrees each way, their face still visible beside it. The screen stays lit and matte with no reflection of the room in it. 17-20s: they lower it slightly, look straight into the lens and land the last line with a small pleased grin, then hold still until the clip cuts. DIALOGUE: the person speaks in their own natural speaking voice, unhurried and conversational, in a normal indoor room, with the mouth movement matching exactly. Across the whole clip they say only this, starting at 0s and finishing by 20s: "Opened, it is the thinnest iPhone ever, with a 7.6-inch inner display that provides a large canvas for viewing content, gaming, and multitasking, and a special nano-texture finish that minimizes glare for an unparalleled viewing experience. And it's in my hands." AUDIO: their voice close on the phone microphone, quiet room tone, the soft mechanical sound of the device opening, and light clothing movement. No music, no narration and no sound effects beyond the room. Only one person and one device exist in this entire video. Keep it subtitle-free: do not generate subtitles, captions, on-screen text, a timestamp, a recording indicator, an interface overlay, a logo or a watermark. No second person, no reflections of another person, no extra hands, no tripod, no studio set and no exaggerated visual effects.
Create a 15-second vertical 9:16 cinematic anime-style food-cooking sequence. Warm hand-painted animation aesthetics, highly detailed backgrounds, clean cel shading, rich textures, natural hand movement, smooth cinematic transitions. Beautifully animated cooking short — not photorealistic, not 3D. No on-screen text, no subtitles, no logos. @product is the breakfast skillet being cooked and served. Same pan, eggs, bacon and finish as the reference, throughout. AUDIO — required: bacon sizzle, egg crack, butter in the pan, spatula scrape, steam. No dialogue, no lyrics, no voiceover. 0-2 seconds Top-down rustic wooden kitchen table. @product sits in the center as a cast-iron skillet surrounded by ingredients: streaky bacon, whole eggs in a bowl, cherry tomatoes, sliced mushrooms, chopped greens, a knob of butter, crusty bread. Warm interior light, cozy traditional kitchen. 2-4 seconds Close-up. A hand lays strips of bacon into a hot black pan over a visible blue gas flame. Fat renders immediately. Bacon curls and sizzles. Oil pops. Camera pushes in slightly. 4-6 seconds Bacon is pushed to the side. Butter melts in the empty half of the pan. A hand cracks two fresh eggs; whites spread, yolks stay bright and centered. Edges bubble. 6-8 seconds Sliced mushrooms and cherry tomatoes hit the pan beside the eggs. They blister and gloss in the bacon fat. Steam rises. This is @product coming together. 8-10 seconds A spatula folds crispy bacon against the eggs. Yolk stays intact. Greens are scattered over the top. Cheese melts into a thin golden veil at the edges. 10-12 seconds Hero close-up three-quarter: the skillet in full — two sunny eggs, caramelised bacon, blistered tomatoes, melted cheese, steam. A hand grinds black pepper over it. 12-13.5 seconds The skillet is lifted off the flame and set onto a wooden board. A slice of toasted bread is placed beside it. Butter gleams. Dynamic settle of the pan. 13.5-15 seconds Hero shot of the finished @product on the wooden table: the skillet breakfast, toast, a small pot of greens, warm golden light. Slow push-in to an appetizing final frame. Visual direction: warm golden lighting, cozy rustic kitchen, anime food animation, hand-painted textures, expressive food movement, smooth transitions, cinematic framing, natural steam and sizzle, consistent hands and kitchen throughout. Negative: photorealism, live action, 3D CGI, plastic textures, malformed hands, extra fingers, deformed food, text, subtitles, logos, watermark.
CHARACTER LOCK — HIGHEST PRIORITY: The person in every shot is exactly the person in @image1. Preserve their facial identity, facial structure, eye shape and spacing, nose, mouth, exact skin tone, hairline, hairstyle, hair length, texture and colour, freckles and distinctive marks, apparent age, build and body proportions — identically, shot to shot. Preserve their gender presentation exactly as shown; do not masculinize, feminize or reinterpret it. Do not beautify, slim, de-age or restyle them. The attached image outranks every word below it: wherever the image and the text disagree, follow the image. Frame 1 is already in the scene — never a studio portrait, grey backdrop, or the attached still as an opening plate. Use they/them throughout. No other people. Create a 30-second late-2000s handheld flip-camera vlog. 16:9. The person in the attached reference is a teenage boy — keep that presentation in every frame. Do not feminize. He is the only person. OPENING — MANDATORY: Frame 1 is already walking toward a tiny beachside meat stand, handheld flip-camera running. Do not open on a studio portrait, grey backdrop, still headshot, or the attached image as a plate. Wardrobe, locked: oversized dusty-blue linen shirt, beige cropped pants, brown flip-flops, canvas shoulder bag. Fit these clothes to the attached person's body. Short boyish hair; wind may move it, do not restyle into long hair. Keep one tiny all-black kitten throughout, same kitten every frame, natural cat behavior. No other people, no vendor face (the stand is self-serve: wrapped cuts on ice, a cash tin, handwritten prices with no readable words). Style: genuine late-2000s handheld flip-camera vlog. Heavy camera shake, focus hunting, accidental reframing, warm faded colors, digital noise, imperfect zooms, no cinematic look. 0-8s — approach: Walking on sand toward the little beachside meat stand. The all-black kitten starts following at his flip-flops. He laughs and points the camera down. Exact lip sync: "Oh... you're coming too?" 8-16s — buy: He picks up a small wrapped cut of meat for himself, drops cash in the tin, then tears off a piece for the kitten. He crouches, smiles, exact lip sync: "This one's for you." The kitten sniffs the meat. 16-24s — feed: On a small wooden stool in the sand beside the stand, he sets the piece down. The kitten eats while he gently pets its head. Exact lip sync: "You were hungry." 24-30s — leave: He walks along the shore carrying the remaining wrapped meat as the same black kitten keeps following beside his flip-flops. He laughs at the camera, exact lip sync: "I have company now." The kitten meows as the recording abruptly ends. Audio: natural beach-stand ambience only — surf, wind, plastic wrap, footsteps on sand, cash tin, kitten meows, the spoken lines. No music, subtitles, logos, or cinematic effects. No on-screen text. Duration: exactly 30 seconds. Aspect ratio: 16:9.
Create a 15-second video that looks like authentic 1890 hand-cranked black-and-white film, filmed by the person holding the camera themselves. **@avatar — THE PERSON.** Preserve their exact face, identity, facial structure, eyes, hair, age, build and proportions throughout. No identity drift, beautification, de-aging or restyling. Ignore the reference background. **WARDROBE:** Wet 1890 London clothing: dark heavy wool overcoat buttoned high, collarless shirt, worn leather boots, soft flat cap, canvas satchel. Damp hair. Outfit never changes. **STYLE:** Authentic early cinema, ~16fps, slightly fast and jerky movement, constant flicker, heavy film grain, dust, hairs, vertical scratches, soft vignette, occasional frame jumps, rain on lens and strong handheld shake. Bad framing is intentional — cropped heads, drifting subject and sudden camera movements. Absolutely no modern cinematic stabilization. **SETTING:** Narrow rainy London street, wet cobblestones, gas lamps, soot-stained brick, market stalls with dripping canvas awnings, horse-drawn carts and pedestrians with umbrellas. **00:00–00:04:** They run through heavy rain toward a market awning, laughing, holding the camera at arm's length as it swings wildly. They shake water from their cap. **00:04–00:08:** They hear something and swing the camera down, searching under a wooden crate. They find a tiny soaked grey-and-white kitten. They crouch down, face filling the frame at an awkward angle. **00:08–00:12:** They dry the kitten with their coat, then buy roasted chestnuts for themselves and a scrap of fish for the kitten. **00:12–00:15:** Sitting on a crate under the awning, they eat together as the rain eases. They laugh into the lens and shrug. The kitten rubs against their boot. The frame stutters, flares white and the film runs out. **AUDIO:** No dialogue, voiceover or music. Only rain, canvas, hooves, cart wheels, distant voices and one small mew. **AVOID:** text, intertitles, subtitles, logos, watermark, color, modern objects/clothing/vehicles/signage, stabilization, cinematic camera work, second person in close-up, face changes or costume changes.
15-second experimental anime fashion film. Use @avatar for the @avatar’s exact identity, outfit and rendering style. A pristine, uniform lemon-yellow cyclorama with a seamless matte floor. Flat, clean yellow throughout, free of grain, gradients and set dressing. No text or logos. The @avatar performs one fast, unconventional dance phrase: ankles weave while the torso stays suspended, a wrist flick ripples through both elbows into a chest pop, knees fold into a low corkscrew pivot, then a heel swivel unwinds the body into an off-balance editorial pose. Continue through asymmetric shoulder rolls, angular arm threading around the face and a sideways glide with the head counter-rotating toward the lens. Every sharp lock immediately releases into another surprising movement. Convincing weight, precise foot contact and expressive follow-through. Electric cyan and hot pink dominate the motion effects. Bold, opaque silhouette afterimages peel away from opposite sides of the body, replay the previous gesture a fraction late and slam back into alignment on each pose. Arm sweeps stretch them into broad curved ribbons; crossing steps weave pink and cyan trails around the ankles; rapid turns fan out overlapping screen-print impressions. At the strongest accents, oversized cyan and pink cutouts of the current pose slide across the foreground, briefly masking the lens. Keep the real character crisp and recognizable, with clean yellow visible between effect bursts. The camera dances in counterpoint: skim backward at ankle height through advancing footwork, whip upward around an outstretched wrist into an extreme face close-up, then use the passing forearm to cut into a rotating overhead view of the low pivot. Dive to a tilted low-angle full-body composition as the character rises. Orbit opposite the sideways glide, exaggerating hands and feet near the wide lens.
A cheerful travel vlogger selfie-hyperlapses through 30 of the world's most iconic landmarks in just 15 seconds. Each location gets exactly one hard-cut shot — half a second, one destination, one gesture (a wave, a peace sign, a thumbs-up) — before jumping straight to the next city and a brand-new outfit. Filmed handheld, selfie-stick style, always keeping her face front and center as the world blurs past behind her, all synced to the beat of the music.
Our First SongRecreate workflow