Grok Imagine Video Prompt - Halloween Parade
A video generation prompt for Grok Imagine featuring characters singing and dancing in a Halloween parade scene.
Esta página reúne exemplos de prompts de Vídeo classificados como “Person Character”. Compare o resultado, a composição ou os planos e a redação original antes de adaptar um exemplo ao seu projeto.
5,763 prompts
Temas
A video generation prompt for Grok Imagine featuring characters singing and dancing in a Halloween parade scene.
Design them dancing and singing the welcome to Halloween song in the Halloween parade
Detailed prompt for testing Seedance 2.5 and Topview T1, covering scenarios like blackboard writing, hand gestures, and anime dance sequences with technical constraints.
①黒板(最初の画1枚・10秒) 朝のカフェ。白紙の黒板に、チョークの線がひとりでに一画ずつ現れ「本日のおすすめ/栗のモンブラン/680円」と横書きで3行。カメラ固定 ②ヒカリ(参照1枚・10秒) 朝の公園。右から歩いてきて正面へ。片手を開いて『5』を見せながら「あと ごふんで はじまるよ!」 ③窓(参照1枚・10秒) 昼の和室。窓へ歩いていき、引き違いの窓を横に押して開ける。レールの抵抗音。風で机の紙がかすかに動く ④アニメ2人(参照1枚・30秒) 2人とも日本の2Dアニメ調。昼の河川敷で、1人がベースを弾き、もう1人が踊ってくるりと回る。ハイタッチして、最後に「おとあそぶ!」。30秒間カットを割らず、カメラは2人のまわりを半周
A detailed cinematic video prompt showing a woman transforming into cyber armor and riding a custom motorcycle.
A cinematic transformation sequence on a busy city market street. A hyper-detailed custom motorcycle labeled “VULCAN” (Limited Edition 1 of 7, Twin Turbo) sits parked, featuring iridescent rainbow exhaust pipes, gold accents, red-and-white tank with a crown logo, exposed mechanics, and glowing blue headlights. A young woman with long dark hair, wearing a black leather jacket and white top, approaches and places her hands on the bike. As she touches it, a smooth, glowing transformation begins: her arms morph into sleek cybernetic limbs with purple energy cores, then the change spreads across her body, replacing her clothing with futuristic metallic armor covered in large faceted purple and blue crystalline lights. Sparks and energy effects erupt as the armor fully forms. She puts on a matching high-tech helmet with a reflective visor, mounts the motorcycle, and rides away down the street. The bike’s exhausts glow with purple thrusters as it accelerates past market stalls, awnings, and pedestrians, leaving a dynamic trail of light and motion blur. Style: highly polished cinematic AI video, realistic lighting mixed with sci-fi neon glows, smooth morphing transitions, dramatic camera movement from close-up to wide tracking shot.
A detailed video prompt for Grok Imagine depicting a cinematic sci-fi scene where armored motorcycles collide and transform into Cybertruck-style vehicles.
Two helmeted riders on massive armored bikes race toward each other on a rain-soaked neon runway at night. They collide in a violent high-speed impact. Immediately after the crash, both bikes begin a dramatic mechanical transformation and fuse together into one heavy Tesla Cybertruck-style vehicle. The newly formed Tesla ignites thrusters and lifts off the runway, flying into the stormy cyberpunk sky. Cinematic photorealistic next-level Transformers sci-fi cyberpunk sequence, ultra-detailed, IMAX-style, 4K. Nighttime on a wet futuristic airport runway under heavy rain, with a glowing neon megacity skyline in the background featuring holographic ads, flying vehicles, and cyberpunk atmosphere. Two massive, heavily armored cyberpunk motorcycles (dark matte black with exposed chrome hydraulics, glowing cyan and magenta energy lines, intricate Transformers-style mechanical detailing) speed toward each other from opposite ends of the runway. Both riders wear full-face high-tech helmets with reflective visors and subtle HUD glow, bodies leaned aggressively forward as tires spray water and engines roar. They collide head-on in a brutal, high-impact crash at the center of the frame — metal crumples, sparks explode, energy arcs crackle, and debris flies in slow motion. Right after the impact, both bikes begin a spectacular mechanical transformation sequence: panels shift and lock, components reconfigure with hydraulic precision, glowing seams light up, and the two wrecked bikes fuse and reassemble into a single massive heavy Tesla Cybertruck-inspired vehicle — angular, industrial, matte black with brushed metal surfaces and red accent lighting. The newly formed Tesla powers up, its undercarriage thrusters ignite with bright blue-white plasma, and it lifts powerfully off the runway. It banks and accelerates forward into the rainy night sky, leaving trails of light, steam, and water spray as the neon city stretches below. Dynamic camera work: low tracking shots following the bikes, dramatic crash close-up with debris flying toward the lens, then a rising aerial pull-back as the Tesla transforms and takes off. Volumetric lighting, heavy rain streaks, lens flares from neon and thrusters, photorealistic metal, water, sparks, and particle effects. Epic high-energy mood with cinematic color grading (deep blues, neon magentas, warm engine glows).
A prompt for creating high-quality Pixar-style 3D animated family comedies, featuring a toddler playing with toy cars and interacting with parents in a cozy bedroom.
Create a high-quality 3D animated family comedy in a warm Pixar-style aesthetic. A cute toddler with big expressive eyes plays with toy cars on a soft bedroom rug while his loving parents watch him affectionately. The baby makes funny facial expressions, becomes stubborn, laughs joyfully, and interacts playfully with his mother and father. Transition to a cozy bedroom scene where the mother rests with the baby, followed by sweet, humorous moments between them. Show the toddler learning to walk, giggling, making adorable expressions, and sharing heartwarming moments with his family. Use cinematic lighting, soft golden sunlight, realistic skin and fabric textures, detailed character animation, expressive eyes, natural body movements, smooth camera transitions, close-up facial shots, medium shots, and wide family shots. Maintain consistent character designs throughout the video. The atmosphere should be wholesome, emotional, funny, and family-friendly, with professional 3D animation, vibrant yet soft colors, and a polished animated-movie look. Vertical 9:16 format, high resolution, smooth motion, and no visual glitches.
A prompt for testing Seedance 2.5 and Kling 4.0 using simple gray-box previz animations, covering office interactions, cycling, and cavalry battles.
共通「動画1は灰色の人形と箱の仮組み。取るのは人の立ち位置・道筋・タイミング・カメラの動きだけ。人形は画像の人物に、箱は本物の場所に置き換える」 01「扉から入った人が、座る人の後ろで止まる→座っていた人が椅子から横にずれて立ち、向き合う。正面の引き→左へ回り込み→2人へ寄る」 02「後ろから自転車が近づき5秒で並び6秒で追い越す→減速して止まり、上半身ごと振り返る。道の同じ側から並走」 03「キャラは最後まで2Dのセル画のまま実写の交差点を渡る。歩行者はちょうど6人。真上の俯瞰→目の高さ」 04「両軍が向かい合って駆け寄り4.6秒ですれ違う→大将が馬を返して刀を掲げる。刀と槍は体に触れず空を切る」
A detailed multi-scene prompt for generating a 34-second hyper-realistic superhero origin story video using Seedance 2.5.
Create a 34-second photorealistic cinematic superhero origin sequence with the look of a high-budget Hollywood movie. A young man in his early 20s with messy dark brown hair, wearing a dark charcoal hoodie, black pants, and sneakers, unexpectedly discovers supernatural powers at sunset. Scene 1 (0–4 seconds): Wide establishing shot of an empty outdoor sports court at golden hour. The young man stands alone against a warm orange sunset. A light breeze moves his hair and hoodie. The camera slowly pushes in, creating a mysterious, emotional atmosphere. Scene 2 (4–8 seconds): He suddenly runs across the concrete court at superhuman speed. Dust explodes beneath his shoes, and the ground cracks from the force of his acceleration. Use dynamic side-tracking shots, realistic motion blur, flying dust particles, and powerful footstep impacts. Scene 3 (8–13 seconds): Close-up of the young man stopping abruptly. He looks at his hands in disbelief as subtle supernatural energy builds. His expression changes from confusion to fear and amazement. Warm sunset rim lighting illuminates his face, with realistic skin texture and shallow depth of field. Scene 4 (13–19 seconds): Match cut to a busy modern city street at golden hour. The young man lands hard in the middle of the road, crouching on one knee. Asphalt fractures beneath his hand, dust and small debris burst outward, and nearby cars brake suddenly. Capture the impact with a low-angle camera and a fast cinematic dolly movement. Scene 5 (19–25 seconds): Intense close-up of his face as he realizes the extent of his powers. He clenches his jaw, breathes heavily, and stares forward with determination. The background streaks with strong directional motion blur while his face remains sharply focused. Use dramatic lighting, realistic expressions, and subtle handheld camera movement. Scene 6 (25–29 seconds): Epic wide shot of the city skyline. The young man launches upward at incredible speed, soaring high above the rooftops. The camera tilts rapidly toward the sky as he becomes a small silhouette against the warm sunset. Convey extraordinary acceleration, realistic scale, and breathtaking cinematic spectacle. Scene 7 (29–34 seconds): Transition to a narrow urban alley surrounded by tall brick walls, metal fire escapes, and scattered trash bins. The young man walks slowly into the alley, exhausted and overwhelmed by what has happened. He stops, looks upward toward the sky, and quietly processes his transformation. End with a dramatic low-angle medium shot and a slow camera push-in. Visual style: Photorealistic live-action, Hollywood superhero origin movie, anamorphic 35mm cinematic lens, 24 fps, HDR, realistic physics, detailed facial expressions, natural skin pores, golden-hour lighting, volumetric dust, atmospheric haze, dramatic composition.
A prompt for generating a hyper-realistic cinematic rom-com video involving a pole vault kiss attempt, using two input images to define character identity.
A hyper-realistic cinematic slapstick romance told across four seamless shots outside a traditional two-story residential building in South Korea during bright afternoon daylight. Natural handheld smartphone cinematography. Use Image Reference 1 as the woman and Image Reference 2 as the man. Preserve both identities, facial features, hairstyles, and clothing consistently throughout every shot. SCENE A modest Korean villa-style apartment building with pale concrete walls, an open second-floor window, and a white bedsheet hanging from an outdoor laundry rack. A narrow pedestrian walkway runs beneath it, surrounded by potted plants, parked scooters, utility cables, and neighboring buildings. SUBJECTS A Korean woman with her hair tied in a neat bun, wearing a casual patterned house dress and simple indoor slippers. A scruffy, eccentric-looking Korean man with messy hair, light stubble, worn casual clothes, and old sneakers. SHOT 1 | THE UNEXPECTED ENCOUNTER The woman stands at the second-floor window, struggling to straighten a large white bedsheet with a long bamboo laundry pole. Suddenly, the pole slips from her hands and falls toward the street. The camera tilts downward, revealing the man walking beneath the window while looking at his phone. The bamboo lands beside him with a loud THUNK, startling him so badly that he stumbles backward and falls onto a soft pile of laundry bags. He looks up in disbelief. The woman leans out of the window, embarrassed. Woman: "Oh my God! I'm so sorry!" Man: "It's okay... but you owe me a kiss!" SHOT 2 | THE FLIRTY CHALLENGE The woman rolls her eyes but notices the man's exaggeratedly charming smile. She laughs despite herself. Woman: "A kiss? You must be dreaming!" The man dramatically places one hand over his heart and pretends to faint. She giggles, covering her mouth, while he suddenly springs back up and points toward the window. Man: "Then let me come up and prove I'm worth it!" She shakes her head, smiling, and gestures for him to wait. SHOT 3 | THE IMPOSSIBLE ROMANTIC MOVE The man grabs the fallen bamboo pole, runs several steps, and plants it firmly against the ground like a pole vaulter. The bamboo bends dramatically as he launches upward toward the second-floor window. The handheld camera races upward, tracking his ridiculous flight with dynamic motion blur. The woman gasps as he arrives at window level, clinging to the flexible bamboo. He steadies himself, leans forward, and puckers his lips comically. The woman initially looks shocked, then starts laughing at his ridiculous expression. Man: "So... can I have my kiss now?" She pauses for a beat, looks into his eyes, and unexpectedly leans forward. SHOT 4 | THE KISS AND THE LAUNCH The woman gives him a quick, playful kiss on the lips. Both freeze for a split second, eyes wide with surprise. The man beams with happiness. Man: "Worth every second!" But the overloaded bamboo snaps, sending him tumbling back down into the laundry bags.
A cinematic fantasy video prompt depicting Dracula in black crystal armor riding a blood dragon through mist under a full moon.
Dracula in Black Crystal Armor Riding a Blood Dragon through the mist during a full moon
A detailed scripted prompt for a 30-second cinematic comedy video about a zombie and an alien fighting over the last donut in a shop.
A zombie and an alien walk into a donut shop at 2am only one donut left FORMAT: 30 seconds | 24fps cinematic | ultra photorealistic live action | studio comedy action | practical prosthetic creature makeup | VFX time freeze | 2am fluorescent night interior. Real-time, no random slow motion. NO ON-SCREEN TEXT - CRITICAL: no captions, subtitles, titles, logos, watermarks, timecode or UI at any point. Menu board and case labels are blank coloured panels, never legible. IDENTITY LOCK: the same two leads in every shot, never changing face, skin, height, wardrobe or props. No morphing, no extra fighters. The cashier and four customers become frozen statues after 0:03 and never move, blink or react until 0:27. CHARACTERS THE ZOMBIE: 6 foot 2, gaunt, grey-green mottled skin with dry cracked patches, cloudy pale eyes, sunken sockets, lank dark hair, no wounds, no blood. Torn navy night-shift security uniform, crooked clip-on tie, one shoe untied. Moves stiff and heavy, then surprisingly quick. Voice: deep gravelly rasp, slow, every word an effort. THE ALIEN: 4 foot 10, slender, smooth pale lilac skin, large glossy black almond eyes, small nostril slits, narrow bald cranium, long four-fingered hands. Oversized yellow tourist rain poncho, small cross-body camera. Voice: light, clipped, overly correct English with a faint electronic flutter. ENVIRONMENT: a narrow 24-hour donut shop at 2am. Cream tile floor, chrome stools, a lit glass display case along the right wall, all trays empty except exactly ONE glazed donut on a paper doily, glowing under the case light. Rain streaking the front window, a red traffic light outside. 0:00-0:03 THE FREEZE Handheld medium down the counter. A customer drops a coffee cup, coffee arcing. The zombie and the alien lunge at the case from opposite ends. Sound drops out. Everything stops: coffee hangs as a glossy brown ribbon, raindrops pinned on the glass, the cashier's yawn locked open. Only the two leads keep moving. 0:03-0:08 REALIZATION Both straighten, look at the hanging coffee, then each other. Slow 30 degree orbit. ALIEN: "I have crossed four galaxies for this donut." ZOMBIE: "I... died... for this donut." ALIEN: "That is not the same thing." 0:08-0:13 FIRST EXCHANGE The zombie swings a long stiff arm, the alien ducks, slides on wet tile and kicks a stool into his shins. He topples, catches the counter, groans. Whip pan as she vaults onto the counter. 0:13-0:18 THE FLASH She snaps her camera in his face, a hard white flash. He staggers, then pulls a folded umbrella from a frozen customer with real effort, the customer staying rigid. ZOMBIE: "Rude." 0:18-0:23 COUNTER FIGHT Fast tracking shot. She leaps stool to stool over frozen shoulders, he swings the umbrella, she parries with a frozen customer's rolled newspaper. Three crisp readable exchanges. His clip-on tie pops off. Her hood flops over her eyes. 0:23-0:27 TRUCE Both stop, breathing hard, hands on knees, coffee ribbon between them. ALIEN: "Share?" ZOMBIE: "Share." They shake hands, four lilac fingers in a grey palm. Micro pause. Both eyes flick to the donut. 0:27-0:30 TIME RESUMES Both dive at the case. The world unfreezes: coffee splashes down, the cashier finishes the yawn, customers jolt, rain runs, fridge hum slams back. LAST FRAME: locked medium wide down the counter. The zombie and the alien pressed cheek to cheek at the open case, both gripping the one donut, glaze stretching between their hands, both grinning, coffee spraying through the light. Cashier and customers turn to stare. No text, captions, subtitles, signage lettering, logos or watermarks anywhere in frame. CAMERA: controlled handheld in exchanges, smooth dolly and orbit on freeze reveals, low angle for the zombie, eye level for the alien, fast whip pans. Shallow depth on dialogue, consistent screen direction. LIGHT AND COLOUR: cold overhead fluorescent key, warm amber case glow on faces and the donut, red street light through the rainy window. 70% cool tile and skin, 20% case amber, 10% red. VFX AND PHYSICS: frozen liquid holds surface tension and highlights. The resume restores gravity in one continuous frame, no morphing, no dissolve. Stools scrape and tip, his heavy fall shakes the counter, she lands light, poncho crackling and lagging, true contact shadows. Nothing floats except the deliberately frozen. ACTING: natural blinking, active brows on the zombie, the alien's eyes narrowing and widening, dry deadpan timing. Only the speaker's mouth forms words. AUDIO, DIEGETIC ONLY: fridge hum and rain before the freeze. During it near-silence: breathing, a dry groan, sneaker squeak on wet tile, poncho crackle, stool scrape, umbrella whoosh, shutter clack. At resume: splash, gasps, rain and hum return. These are the only words spoken. No music, no score, no humming, no swell, no drone. LOCKS: No blood, no gore, no wounds. No duplicated people, no extra limbs, the alien always has four fingers. No wardrobe or face changes, no cartoon physics, no excessive shake. Dry comedy, readable fight, coherent physics.
A narrative prompt describing a meeting between Elon Musk, Optimus Prime, and a fantasy ocean queen on Mars.
POV: Elon and Optimus meet Queen Thalassa, ruler of Nereia, the hidden ocean city beneath Mars. 🌊🔴 Elon made first contact. Optimus got his first pet. 🐟🤖
A complex prompt for a 30-second Halloween folk horror video featuring giant monsters and heroes using Seedance 2.5 with consistent image references.
THE HARVEST KING — SEEDANCE 2.5 DIRECTIVE: Exactly 30 seconds, 16:9, photorealistic Hollywood folk-horror, 60fps where supported. Preserve @image1’s exact face, hair, skin tone, build, outfit and props in every shot. No identity drift, clones or costume changes. LOCATION: Haunted pumpkin farm on Halloween night, foggy field, corn maze on the left, weathered red barn far right, rusted tractor, blue moonlight and glowing orange pumpkins. Maintain geography. MONSTER: A 15-meter Harvest King made of vines, rotten pumpkins, tractor wheels and farm tools. Giant carved pumpkin head, burning candle inside its mouth (weak point), enormous rusted scythe. Supporting enemies: burlap scarecrows with ember eyes, stitched mouths and rusted sickles. WEAPON: @image1’s existing equipment plus one rusted iron lantern-hook pole pulled from a fence. Grounded, weighty combat and realistic physics. TIMELINE: 00:00–00:05 | THE AWAKENING: Macro jack-o’-lantern candle sputters. Crane shot reveals the foggy field as orange light ignites pumpkins toward the barn. @image1 turns toward the glow, alert. 00:05–00:10 | SCARECROWS: Scarecrows break free and attack. The hero tears a rusted iron pole from the fence, dodges a sickle and strikes an attacker into the others. Track laterally through corn. 00:10–00:15 | CROW STORM: Five scarecrows rush him. He blocks sickles, hooks a leg and counters. One powerful boot strike shatters a scarecrow into straw and exploding crows during a 180° slow-motion orbit. Music abruptly fades. 00:15–00:20 | THE GIANT RISES: Earth splits; pumpkins roll inward. The colossal Harvest King erupts from the soil, towering over the barn. It swings its scythe; the hero slides underneath as the blade splinters the fence and barn wall permanently. 00:20–00:25 | THE CLIMB: The hero sprints, leaps onto the embedded scythe handle and climbs onto the giant’s arm. He hooks whipping vines with his pole, swings past them and reaches the shoulder, facing the burning candle inside its mouth. 00:25–00:30 | FINAL TWIST: Flame-lit close-up. He drives the pole through the carved mouth, smashing the candle. A fiery blast erupts; he leaps clear. The giant collapses into rotten pumpkins and straw in slow motion. At dawn, the hero rises among motionless scarecrows. At 00:29.5, the untouched jack-o’-lantern relights and its smile widens. Cut to black at exactly 00:30. AUDIO: Original score: eerie cello and music box, crows, wind, creaking wood, taiko and choir escalating into 140-BPM orchestral action. Synchronize impacts, metallic scythe whistle, flame blast and giant collapse. Absolute silence at 00:28.5; one final music-box note. No dialogue, text or logos. Maintain continuity of damage, fire, debris and character identity.
Detailed prompt for generating a 15-second vertical comedy video focusing on a woman in a gym, emphasizing realistic mobile phone cinematography and natural lighting.
15秒超写实搞笑短视频,9:16竖屏,手机原相机纪实质感,真实健身房环境,柔和室内顶光,全程像朋友在旁边随手跟着拍,轻微自然晃动,生活化非常强。 【人物设定】 高颜值年轻亚洲女生,自然淡妆,五官精致舒展,身材匀称,气质温柔清冷,真实路人感,皮肤保留自然纹理和细微汗珠,不网红磨皮,不夸张整容脸,身体比例自然。面部和身材完全保持与参考图一致 镜头1|0-4秒 假象自律 明亮干净的健身房,背景有少量正在训练的人影和器械。女生在自由训练区认真做标准深蹲,动作流畅稳定。镜头从画面右后方手持跟拍,先轻缓后退保持人物全身入镜,再随着她起身微微上抬,像朋友边走边拍,画面有轻微呼吸感和自然晃动,营造她非常自律刻苦的错觉。 镜头2|4-8秒 误导继续训练 硬切到侧面跟拍,女生站立举小哑铃做弯举训练,微微出汗,认真发力,表情冷静专注。镜头贴近她身体侧前方跟随,先轻微横移,再缓慢推近到上半身,背景器械自然掠过,观众以为这是一条自律健身视频。 镜头3|8-12秒 第一层反转 女生突然放下哑铃,表情从专注变成松弛,长舒一口气。镜头继续手持跟随她走到训练凳旁。她坐下,从随身健身包里掏出炸鸡和冰奶茶,动作自然熟练,撕开包装袋。镜头在她掏东西时快速但轻微地推近,形成抓拍式喜剧效果。 镜头4|12-15秒 终极反转 镜头直接推到近距离人脸和炸鸡奶茶同框。女生坦然自然地咬一口炸鸡,再喝一口奶茶,看向镜头,用轻松随意的语气说:“吃饱了才有力气减肥。”说完表情淡定,像在说一件很正常的事。结尾停留在她淡定干饭的反差脸上,手持镜头有极轻微的前后晃动感,强化朋友偷拍感。 【画面质感】 超写实真人实拍,原相机高清画质,真实皮肤纹理、毛孔、汗珠和自然肤色,健身房轻微环境噪点,室内顶光自然,不要电影级精致布光。镜头运动以手持跟拍、轻推近、轻横移为主,节奏随性,不做稳定器丝滑运镜。场景之间直接硬切,不做花哨转场。 【音效】 健身房远处器械碰撞声、轻微脚步声、女生训练呼吸声、包装袋撕开声、炸鸡咀嚼声、奶茶吸管声。台词清晰自然,不要夸张配音感。台词说完后加入短促轻松搞笑音乐点。 【负面提示词】 畸形手指,扭曲肢体,多指,少指,脸部崩坏,五官错位,AI假脸,塑料皮肤,过度美颜,硅胶脸,夸张美瞳,变形身体,比例失调,头身比例异常,悬浮物体,物品穿模,画面闪烁,物体漂移,卡通渲染,3D动画,手绘,插画,屏幕字幕,文字,花字,水印,logo,边框,强烈滤镜,柔光磨皮,稳定器顺滑运镜,固定机位,僵硬摆拍,舞台打光,惨白假光,过度布光,多余人物,杂乱抢镜杂物,夸张表演,尬演,做作表情
Detailed prompt for generating a high-end editorial style Halloween fashion and beauty commercial using Seedance 2.5.
**FORMAT:** 15s · 16:9 **CONCEPT:** THE NIGHT BELONGS TO YOU **STYLE:** Ultra-fancy Halloween fashion and beauty campaign, high-end editorial cinematography, extravagant styling, rich textures, dramatic lighting, rapid cinematic transitions. **0–3s | THE REVEAL** A grand gothic dressing room bursts into view, overflowing with extravagant Halloween costume props: feathered witch hats, jeweled masquerade masks, black lace gloves, vampire capes, ornate crowns, crystal wands, colored contact lenses, theatrical makeup palettes, and elaborate wigs. A stunning adult blonde model in a luxurious black-and-orange gothic costume turns toward the camera. A dramatic flash of light reveals the entire collection. **3–6s | COSTUME FANTASY** Rapid, seamless transformations: the model becomes a glamorous witch with a sculptural black hat, a gothic vampire in a velvet cape, and a mysterious masked queen wearing an elaborate jeweled headpiece. Each look is extravagant, fashionable, and visually distinct. **6–10s | BEAUTY CLOSE-UPS** Extreme close-ups of glittering eyeshadow, glossy dark-red lips, rhinestone face gems, intricate lace masks, and sparkling accessories. The model applies lipstick, slips on jeweled gloves, and lifts an ornate mask toward her face. Rich orange, burgundy, black, and amethyst tones. **10–13s | THE GRAND FINALE** The model stands center-frame in a breathtaking Halloween couture look, surrounded by an extravagant display of costumes, wigs, masks, pumpkins, and beauty products. Fabric sweeps dramatically through the frame as the camera pushes in. **13–15s | CAMPAIGN END FRAME** A luxurious flat lay of the complete Halloween collection, with makeup, costume accessories, masks, and jeweled props arranged like a fashion editorial. Bold elegant typography appears: **“OWN THE NIGHT.”** **“HALLOWEEN COLLECTION”** **VISUAL DIRECTION:** Every frame should feel lavish, colorful, and editorial. Abundant Halloween props with deliberate art direction, intricate costume details, glossy beauty lighting, cinematic camera movement, and seamless transitions. Premium fashion-house campaign meets extravagant Halloween fantasy. No cheap party-store aesthetic, no cluttered composition, no cartoon effects, no low-quality CGI. No voiceover or background music.
This prompt tests the optical rack focus function, transitioning focus from a foreground model to a background silhouette.
A high-tension optical rack focus with hard-flash illumination. The frame starts with shallow focus locked squarely on the sculptural kiss-curl and sharp eyeliner of the female model in the foreground. On beat, the focus violently racks to the background wall, bringing the razor-sharp edge of the elongated silhouette shadow and the male model's direct stare into crisp focus, throwing the foreground into painterly softness.
A prompt for a 10-second hyper-realistic video showing a woman driving in 2050 with flying cars overhead and detailed character descriptions.
TITLE: POV — DRIVING IN 2050 Create a 10-second ultra-photorealistic futuristic video showing a beautiful young woman casually driving along a highway in the year 2050, while hundreds of autonomous flying vehicles travel overhead in perfectly organized aerial traffic lanes. MAIN CHARACTER — STRICT IDENTITY CONSISTENCY A beautiful 23-year-old woman with long, straight pastel-pink hair, delicate facial features, realistic fair skin with visible natural texture, soft pink makeup, and subtle glossy lips. She wears: A fitted blue denim sleeveless top. A futuristic metallic silver choker necklace. Bright blue statement earrings. A black seatbelt across her chest. Maintain her exact face, hairstyle, outfit, proportions, and accessories throughout. SCENE 1 — 0:00–0:02 | INSIDE THE CAR Start with a fixed dashboard-mounted smartphone camera facing the woman from the passenger side. She sits behind the steering wheel of a modern luxury vehicle, driving casually. Warm golden-hour sunlight streams through the side window, illuminating her pink hair and face. She looks ahead at the road, then briefly glances toward the camera with a subtle, amused smile. The vehicle interior is realistic, with black leather seats, a dark roof, and natural windshield reflections. SCENE 2 — 0:02–0:04 | THE FUTURISTIC HIGHWAY REVEAL Hard cut to a spectacular elevated wide-angle view of a massive multilane highway at sunset. A dark luxury SUV drives along the nearly empty highway. Above the highway, hundreds of futuristic autonomous flying vehicles hover and travel in tightly organized aerial traffic lanes. The flying vehicles have compact aerodynamic bodies, metallic silver and dark-gray surfaces, and advanced concealed propulsion systems. They move smoothly in synchronized formations, maintaining precise distances from one another. Some fly directly above the highway while others stretch toward the horizon, creating a dense futuristic transportation network. Golden sunlight reflects off their metallic surfaces. SCENE 3 — 0:04–0:06 | BACK TO THE DRIVER Hard cut back to the same dashboard camera angle. The pink-haired woman continues driving completely unfazed by the extraordinary traffic system outside. She rests one hand naturally on the steering wheel and looks toward the windshield with a playful, slightly impressed expression. Sunlight shifts across her face as the vehicle moves. Her facial movements are subtle, natural, and believable. SCENE 4 — 0:06–0:08 | MASSIVE AERIAL TRAFFIC Cut to a dramatic elevated tracking shot following the dark SUV along the highway. The camera reveals an enormous formation of flying vehicles stretching across the sky. Several vehicles glide directly above the camera while hundreds more move across multiple aerial levels. The futuristic traffic looks technologically advanced but physically believable. Below, the dark SUV continues along the sunlit road, appearing calm amidst the chaos above.
A detailed horror cyberpunk video prompt for Grok Imagine, showing a suitcase transforming into a bio-mechanical dragon on a rainy city rooftop.
Cinematic realistic video prompt (horror cyberpunk Transformers evolution): Nighttime cyberpunk megacity rooftop, rain-slicked concrete and neon reflections. A exhausted, rain-soaked man in a worn dark trench coat stands at the edge of a towering building. He hurls a heavy metallic suitcase off the roof with desperate force. The suitcase tumbles through the rainy night sky. Mid-fall, the suitcase violently transforms with mechanical horror: panels rip open, hydraulic limbs explode outward, gears and cables tear through metal skin with wet, grinding, flesh-like sounds. It rapidly unfolds and mutates into a massive biomechanical dragon — half machine, half nightmare. Segmented metal scales shift like living armor, glowing red cybernetic eyes snap open, wings of jagged steel and translucent membrane unfurl with bone-cracking force, and a long neck snaps forward with a maw full of interlocking blades. Sparks, black fluid, and neon-lit steam erupt during the transformation. The dragon fully forms mid-air, lets out a distorted, mechanical roar that mixes with distorted city alarms, then dives or circles with predatory menace. Dark, wet, industrial atmosphere. Extreme detail on metal textures, rain, reflections, and the grotesque beauty of the transformation. Shot with cinematic anamorphic lens, slow-motion emphasis on the transformation, dynamic tracking camera that follows the suitcase as it falls and becomes the dragon. High-contrast neon lighting (cyan, magenta, deep red), volumetric rain, volumetric god rays from neon signs, realistic physics, 8K, photorealistic, horror-sci-fi tone, next-level Transformers-style evolution with terrifying organic-mechanical fusion.
A complex multi-shot prompt for Seedance 2.5 featuring a woman flying a car to escape Manhattan traffic, only to find the sky equally crowded.
Image1 is CAR: a white wedge coupe, black glass canopy, red full-width tail light bar, rear wing raised on two struts, four round thruster nozzles under the rear bumper. Image2 is NORA: long straight blonde hair, light freckles, white ribbed tank top. Image3 is INTERIOR: the cabin of CAR, white quilted bucket seats, dark grey suede dashboard, a yoke and a wide centre screen. NORA, stuck in Manhattan gridlock in CAR, presses a button and lifts straight up out of the traffic, then finds the sky just as jammed. 0-3s: Fast forward-moving shot just above the car roofs down East 42nd Street: chaotic gridlock of sedans, SUVs, vans, a box truck, a few yellow taxis, motorcycles and cyclists weaving between lanes, people crossing between bumpers, pushing in on CAR stuck in the middle lane beside a city bus. <horns> <idling engines> <a bike bell> 3-6s: Medium side shot from the passenger seat in Image3: NORA at the yoke in profile, the bus filling the window beyond; NORA taps a thumb on the yoke, breathes out through the nose and glances down at the centre screen. <horns outside> 6-7s: Close on the centre screen: a finger of NORA presses a glowing round button. <a soft chime> 7-10s: Low at road level behind CAR beside the bus: the four nozzles flare white, gas blasts onto the asphalt, a cyclist swerves, and CAR rises straight up past the side of the bus, level, the Chrysler Building ahead. <thrusters spooling up into a roar> 10-13s: From the back seat over the shoulder of NORA, through the windscreen: CAR already above the bus roof, the jammed street spreading out below, the Chrysler Building ahead. <steady thruster roar> 13-15s: Close front on NORA in the cabin: a small pleased half smile that fades as NORA looks ahead. 15-19s: NORA's view through the windscreen at the 20th floor: dozens of flying cars between the towers, first gliding in neat stacked lanes, then several break out and dart everywhere, cutting across lanes, diving and swerving, thrusters glowing. <thrusters humming> <horns in the air> 19-22s: Medium front on NORA in the cabin: NORA shuts the eyes, lets out a tired breath, rubs the forehead and sinks back into the seat. <a tired sigh> Constraints: hard cut on every timestamp. CAR keeps the body of Image1 whole, the nozzles dark until 7s; NORA matches Image2 and Image3 is the only cabin. Natural understated acting, small real expressions, no mugging. No dialogue. Raw unedited RED camera footage on a real New York street: very shaky handheld camera in every shot, flat overcast daylight, low contrast, low saturation, no grade, real skin. A faint low drone under the sound, no other music. Plates and signs blank, no readable text, no logos, no watermark.
A detailed prompt for a 30-second cinematic video showing a female acrobat performing on wooden poles over a turquoise lake, emphasizing realistic physics and camera movement.
Create it 30-second video featuring a breathtaking cinematic acrobatic performance above a beautiful turquoise lake surrounded by lush green mountains, misty cliffs, and soft cloudy skies. A graceful female acrobat wearing a flowing white outfit performs incredible flips, handstands, aerial rotations, and perfectly controlled balancing moves on a series of tall, narrow wooden poles rising above the water. The camera captures her movements from wide cinematic angles and smooth tracking shots, highlighting her strength, flexibility, and balance against the stunning natural landscape. She leaps gracefully from one pole to another, spins through the air, and lands with precision on a wooden platform beside the lake, finishing with an elegant martial-arts-inspired pose with her arms extended. Use realistic human anatomy, natural body movements, accurate physics, fluid transitions, soft daylight, atmospheric mountain fog, crystal-clear water reflections, and ultra-realistic cinematic details. Keep the acrobat's face, hairstyle, outfit, and body proportions consistent throughout the video. The final result should look like a professionally filmed real-life acrobatic performance, with smooth camera movement, breathtaking scenery, and no text, logos, or visual distortions.
A detailed prompt for generating a realistic 30-second cinematic historical scene set in 1950s Korea, depicting a young woman's daily chores.
Create a highly realistic 30-second cinematic historical sequence set in Korea during the 1950s. The entire video should feel like a naturally observed fragment of ordinary life from that era — not a vlog, not a modern recording, and not a person performing for a camera. MAIN SUBJECT: A young Korean woman in her early 20s with a consistent, natural appearance and believable facial features. She has black hair tied loosely back, minimal makeup, natural skin texture and subtle everyday expressions. Her outfit is period-inspired but youthful and attractive: a fitted cropped short-sleeve cotton top paired with high-waisted shorts ending around the upper thigh, simple flat leather shoes, and a small understated necklace. The clothing should look like authentic, practical 1950s cotton clothing rather than modern fashion. Natural fabric wrinkles, slightly faded colors and realistic stitching. CONCEPT: Show several small moments from her ordinary daily routine, progressing naturally from inside her home to the neighborhood outside. She never looks into or acknowledges the camera. There is no visible camera, filming equipment, phone or modern recording device. SCENE 1 — MORNING AT HOME | 0–6 SEC Inside a modest 1950s Korean home with wooden floors, simple bedding, handmade household objects and sunlight entering through a paper-covered window. The woman folds a small stack of freshly washed clothes, then carries them outside in a woven basket. Her movements are casual and unhurried. Sunlight falls naturally across the room, revealing dust particles in the air. SCENE 2 — HANGING LAUNDRY | 6–11 SEC She steps into a small courtyard behind the house and hangs clothes on a simple rope line. Show her stretching upward to place a shirt over the line, securing it with simple wooden clothespins, then smoothing another garment with her hands. A light breeze gently moves the hanging clothes and her loose hair. Include small environmental details: clay pots, wooden walls, stone ground, a water basin and neighboring rooftops in the background. SCENE 3 — PREPARING TO GO OUT | 11–16 SEC She returns inside and quickly prepares to leave. She picks up a small woven shoulder bag, checks herself naturally in a small old-fashioned mirror, adjusts her hair and puts on her simple shoes. Her expression is relaxed and spontaneous — never posed. SCENE 4 — WALKING THROUGH THE NEIGHBORHOOD | 16–22 SEC She walks out through a wooden gate into a quiet 1950s Korean residential lane. Show traditional houses, stone walls, bicycles, wooden carts, laundry lines, children playing farther away and neighbors performing ordinary chores. She walks naturally through the scene carrying her small bag. SCENE 5 — EVERYDAY ERRAND | 22–27 SEC She approaches a small public water pump or communal washing area. She briefly fills a small metal container, adjusts her grip and continues walking. Keep the interaction simple and believable. No vendor, no market performance and no staged interaction. SCENE 6 — FINAL MOMENT | 27–30 SEC She walks along a sunlit lane bordered by old Korean houses and trees. A breeze moves her clothes naturally as she passes through patches of sunlight and shadow. She briefly pauses to look toward something happening farther down the street, smiles subtly, then continues walking out of frame. VISUAL STYLE: Authentic 1950s Korean period-film aesthetic. Photorealistic historical reconstruction with soft natural daylight, slightly faded monochrome or extremely subtle muted color, fine analog film grain, gentle halation, slight lens softness, natural exposure fluctuations and occasional tiny imperfections characteristic of period film. The image must NOT look like a modern video with a generic "vintage filter." Architecture, clothing, props, materials and human behavior should all convincingly belong to 1950s Korea. CAMERA: Observational period cinematography. The camera exists as an unseen observer rather than a participant. Use carefully composed static shots mixed with very slow natural camera movement. No handheld vlog movement, no selfie perspective, no modern stabilization, no dramatic zooms and no shots where the subject appears aware of being filmed. AUDIO: Pure environmental sound: birds, distant voices, footsteps, fabric moving in the breeze, wooden doors, water flowing from the pump, distant bicycles and neighborhood activity. No music, no narration, no modern sound effects. REALISM: Every movement must obey real-world physics. Natural hands and fingers, realistic walking, believable fabric deformation, consistent shadows, physically accurate sunlight, authentic facial expressions and continuous character identity. No extra fingers, warped hands, floating objects, distorted faces, modern objects or AI-looking motion. The final result should feel like a genuine 30-second fragment of everyday Korean life from the 1950s — intimate, quiet, observational and historically believable, with the woman's ordinary routine providing the visual story. FORMAT: 30 seconds, Seedance 2.5, 16:9, ultra-high-quality photorealism, cinematic historical realism, consistent character identity, coherent transitions, natural motion and authentic 1950s Korean atmosphere.
A detailed prompt for generating a cinematic video of a steam locomotive departure using Seedance 2.5, including steam and camera angle effects.
山あいの小さな機関区。黒い蒸気機関車の足元から、白い蒸気が這ってくる。運転室の窓から見上げるさくら。おじいさんが焚き口に石炭をくべる。レバーを一気に引くと、動輪がその場で空回りし、煙が爆発するように噴き上がる。運転室ごと真っ白。蒸気が晴れると、すすだらけの顔で咳き込むさくら。おじいさんが吹き出し、軍手の手をそっと重ねて、レバーをほんの少しだけ戻してみせる。 すすの残った顔で、レバーをじわっと少しだけ開ける。煙突がぽっ、ぽっと鳴り、動輪がレールをつかんで回り出す。窓から身を乗り出すと、髪が風に流れる。汽笛のひもを引くと、白い蒸気が真上に噴き上がる。線路わきの白い小旗が近づく。運転室でハイタッチ。すすだらけのまま笑う。 一緒に来た友達が手持ちで撮った記録。1カット2秒前後、セリフ・字幕・BGMなし。画面に出るのは二人だけ。機関車1両、レバー1本、汽笛のひも1本。番号板・車体・服は文字なしの無地。機関車の形と構内の配置、運転室の中は、人のいない参照画像で固定。
A detailed prompt for a cinematic luxury-style video set in Manhattan, featuring a conversation sparked by a coffee cup mix-up.
The Wrong Coffee — Midtown Manhattan, New York Duration: 30 seconds | Aspect Ratio: 16:9 | Style: Cinematic luxury editorial Character References: Use the two supplied images strictly as character references, not starting frames. The man, Alex, is a confident, successful Manhattan finance executive. Preserve his exact face, dark tailored suit, tie, and hairstyle. The woman, Allisa, is the second character; preserve her exact identity, hairstyle, and elegant fitted outfit. Visual Continuity: Warm, slightly vintage editorial photography with amber highlights, muted contrast, subtle film grain, and realistic skin texture. Consistent color across every scene. No artificial glow or plastic skin. Environmental Accuracy: All visible signage, menus, names, and street text must be correctly spelled and legible. No gibberish. Scene 1 — Morning Coffee (0–4s) Inside an upscale Midtown Manhattan coffee shop with polished stone counters, sophisticated interiors, and busy professionals. Alex approaches the counter, composed and confident. No coffee cups are visible on the counter. Camera positioned behind Alex, looking toward the counter. A beautiful barista in her very early twenties, with striking model-like features and a naturally charismatic presence, turns slightly toward the off-camera coffee preparation area. Alex, casually: "Double espresso, please. My name is Alex." Barista, friendly: "Coming right up." Display exactly "Midtown Manhattan, New York" in the lower third, horizontally centered at the bottom of the frame, Poppins Semi-Bold 600, white (#FFFFFF), on a rounded charcoal-slate (#17212B) background at 60% opacity. No extra characters. Scene 2 — The Barista (4–7s) Camera positioned behind Alex, but a different POV from Scene 1, looking toward the counter. Barista, calling clearly: "Double espresso for Alex!" Alex remains absorbed in his phone. No coffee cups are shown in this shot. Scene 3 — The Wrong Cup (7–11s) Two takeaway coffees are placed side by side on the pickup counter by an unseen employee. One cup is clearly labeled ALEX, the other ALYSSA. Still looking at his phone, Alex reaches with his free hand and mistakenly picks up only the cup labeled ALYSSA. He walks away carrying it, unaware of the mistake. His actual coffee, labeled ALEX, remains on the counter. Scene 4 — The Sidewalk (11–14s) Outside, Alex walks briskly along a busy Midtown Manhattan sidewalk, carrying ALYSSA's coffee in one hand while continuing to look at his phone in the other. Behind him, ALYSSA exits the same café that Alex just came out of, holding his coffee and spots him ahead. Scene 5 — The Encounter (14–19s) ALYSSA catches up to Alex on the sidewalk. ALYSSA, lightly amused: "I think you grabbed my coffee." Alex looks up from his phone, stops, and glances at the cup. Alex, looking at his cup, and then ALYSSA, with an apologetic smile: "Oh, you're right. Sorry about that." They naturally exchange cups. Alex puts his phone away. Scene 6 — Same Direction (19–24s) Alex gestures down the sidewalk. Alex, casually: "Which way are you headed?" ALYSSA, pointing ahead: "Same way as you, apparently." Alex, smiling: "Well, that's convenient." They share a laugh and begin walking together, talking comfortably. A fixed camera captures them moving away down the sidewalk until they exit the frame. Scene 7 — An Unexpected Connection (24–30s) Hard cut to a fixed camera position farther down the same Midtown block. Alex and ALYSSA walk side by side, coffees in hand, enjoying an animated conversation. Allisa, playfully: "So, is this your usual morning routine?" Alex, amused: "Not usually this interesting." Allisa laughs naturally, completing her laugh without interruption. They continue walking and talking, moving toward the right side of the frame. Both characters fully exit the frame to the right. Hold the empty sidewalk for the final 0.5 seconds before ending. Direction: Natural conversational dialogue, accurate lip-sync, subtle emotional expressions, and realistic physical interactions. Alex keeps his phone in one hand throughout Scenes 2–4 and picks up only Allisa's coffee with his free hand. The ALEX cup stays behind until Allisa retrieves it. Maintain exact cup ownership and continuity. Varied cinematic tracking and close-ups, with fixed camera positions for the walking exits. Clean instantaneous hard cuts. Allow dialogue, reactions, and laughter to finish naturally. No slow motion, morphing, subtitles, or music.
A complex prompt for generating a comedy video featuring a grandmother, a leopard, and rent-collecting gangsters with strict identity locking.
REFERENCES: Grandma = 95–100-year-old tiny hunched Korean grandmother (exact face, white bun with wooden pin, quilted indigo jeogori, grey knit scarf, plum skirt, gomusin shoes). Thug = very chubby greasy gangster (exact face with eyebrow scar, shiny black puffer jacket open, red-black floral silk shirt, gold chain, gold ring, black track pants, dirty white sneakers). Leopard = giant adult snow leopard (exact smoky grey-white coat, black rosettes, pale icy-blue eyes, long thick ringed tail). Hut = snowy thatched mountain tea hut on a hilltop with a small snowy yard, a footpath winding down the slope toward camera, and vast misty snow mountains on the right. Keep every identity, outfit, marking and set identical to refs throughout. Only these three beings exist in the scene. PROP LOCKS: PROP-STICK = gnarled dark-brown wooden walking stick, always in Grandma's right hand, she leans on it every moment. PROP-BOWL = small round brown earthenware bowl of raw red meat chunks, always in Grandma's left hand, never dropped, never changes. STAGING LOCK: Grandma stands in the snowy yard just in front of the Hut doorway, facing the footpath/camera, for the entire scene; she never walks away. Thug comes UP the footpath and stops 2 metres in front of her, his BACK to the footpath and slope. Leopard later comes up that SAME footpath from the slope behind Thug. Snow falling softly throughout. LOOK: overcast winter daylight, cool blue-white snow, warm amber glow from the hut door, low saturation, low-to-mid contrast, filmic grain, soft halation. Ambient wind, crunching snow. PACING: natural unhurried Korean delivery, short pauses between lines; never rushed. 0–6s: Opening aerial FPV drone shot: high above vast misty snow mountain ranges, gliding smoothly over frosted pine ridges and drifting clouds, then diving gracefully down the slope toward the tiny snow-covered Hut on the hilltop, slowing and pushing in to a medium shot of Grandma standing in the yard, leaning on PROP-STICK, holding PROP-BOWL. 6–11s: Grandma looks around the hills, calling sweetly and caringly, like to her own son: Grandma (sing-song, tender): 「우리 아가~ 밥 먹자… 어디 갔니, 우리 아들~」 She calls once more, peering at the ridges, smiling. 11–17s: Heavy crunching steps. Thug stomps up the footpath, panting, chin raised, and stops in front of her. Thug (rough, ruthless): 「할매, 이번 달 자릿세. 빨리 내놔.」 Grandma looks at him calmly and gives a small sly smirk. Grandma (humble, gentle): 「잠깐만… 우리 고양이 밥 좀 주고. 그다음에 줄게.」 17–22s: Thug scoffs proudly, pinching his thumb and index finger close together in front of his face to mock something tiny. Thug (angry, sarcastic): 「그 쪼끄만 망할 고양이 기다릴 시간 없어!」 Grandma's eyes shift past his shoulder to the slope; a happy, satisfied smile spreads. Grandma (warm): 「아이고… 저기 오네.」 22–30s: Thug turns around just as the giant snow leopard lunges from the slope behind him.
A detailed prompt for Seedance 2.5 depicting a live-action style fantasy scene of a rural woman making popcorn with a small dragon.
Live-action cinematic fantasy scene, 18 seconds, horizontal 16:9, one continuous story without time skips. A young village woman leans over a cast-iron pot of dry corn on an outdoor table when her tiny pink baby dragon LEAPS UP onto the table and breathes ONE jet of fire into the open pot, blowing popcorn all over her face. After a short scolding the dragon learns to heat the pot gently by lighting the firewood under it, and in the last seconds the lid is lifted on a pot of perfect, unburnt popcorn: she says "Perfect... you did it!" and both of them smile. REFERENCES Reference image 1 = the woman, reference image 2 = the baby dragon (the standing baby dragon with huge amber eyes). Match their faces, hair, colors, costume, scales, horns and wings exactly in every shot. MAIN SUBJECT Woman: about 20 years old, light brown (chestnut) hair in a messy loose high bun with wispy loose strands around the face, light grey-blue eyes, freckles across the nose and cheeks, warm natural smile. Cream textured linen blouse with a gathered round neckline and full sleeves rolled up to the forearms, dark charcoal-grey wool pinafore dress with wide straps, long dark wool skirt, simple brown leather boots, a folded linen cloth tucked into the pinafore. She speaks English in a natural, warm voice. Baby dragon: a SMALL, slim, lizard-like baby dragon with a big head, the size of a large cat, about 45 cm long and 25 cm tall at the shoulders. Cream-white pebbled scales with soft pink and lilac mottled patches on the back, legs and tail, a smooth cream scaled chest and neck. A tall fan of pink-tipped spikes (pink to crimson tips, lilac bases) sweeps up and back from the top of the head like a crest, with smaller cream spikes around the cheeks and neck. TWO cream ridged horns curving up from the head, the right one taller than the left. VERY LARGE round glossy amber eyes with black vertical pupils, pink skin around the eyes, a small pale pink nose, a gentle closed-mouth smile. One large translucent pale wing visible with pink ribs and fine pink veins, wingspan about 70 cm, and a wing arm covered in scales. A tail that curls upward, ending in a fan of pink-tipped spikes. Four sturdy legs with cream claws. Quick, curious and playful like a puppy. It does not speak, only makes tiny squeaks and chirps. LOCATION AND HEIGHT Stone courtyard of a rustic cottage at ground level. Rough wooden table 2 m long and 0.9 m high. In the middle of the table lies a flat grey stone slab, 60 cm wide and 5 cm thick. On the slab stands a small black iron tripod, 25 cm high, and under it lies a neat stack of split dry firewood logs with a handful of dry kindling and twigs between them. The firewood is UNLIT from the start: no flame, no embers, no smoke. There is NO other fire source anywhere in the scene. Foreground: on the table, a black cast-iron pot 35 cm wide with a domed lid with a knob, filled with dry corn kernels (at the start it stands on the bare slab beside the tripod), a wooden bowl of kernels, loose kernels on the planks. Midground: a mossy dry-stone wall 1.3 m high, a grey stone cottage on the left with a dark wooden door, a stack of firewood against the cottage wall. Background: green hills and sea cliffs about 3 km away under a golden sunset sky. Nobody else is present. CAMERA AND LOOK Handheld cinematic live-action, horizontal. Shot sizes: medium-wide, three-quarter, close-up. One low warm sun from the left behind the wall, golden hour, soft long shadows, natural skin tones, muted contrast. Mild bloom, faint film noise, slight hand shake, delayed autofocus right after the fire burst. The only extra light is the small orange glow of the dragon's flame. No glossy beauty lighting, no artificial rim light, no heavy grading. The camera and the operator never appear. TIMELINE (CONTINUOUS STORY) Cut only after a movement is finished. The dragon is always on the woman's RIGHT side, on the table. The stone slab with the tripod and firewood stays in the middle of the table. 00:00-00:03.5 THE LEAP AND THE FIRE (one continuous medium-wide shot, slow push-in) The first frame is already in motion. The woman leans over the open pot on the table, both hands on the table edge, looking down at the corn. The unlit firewood and tripod are visible on the slab beside the pot. 00:00-00:00.5 TAKEOFF: at the right edge of the frame the baby dragon pushes off the dirt ground with its hind legs, wing snapping half open. 00:00.5-00:01.3 FLIGHT: it arcs about 1 m up and across the table edge in a short rising curve, wings fluttering. 00:01.3-00:01.8 LANDING: its front claws hit the wooden planks first, then the hind legs, it slides about 10 cm and scatters a few kernels, the tail whips for balance and the wings fold. The woman freezes, eyes wide, still leaning over the pot. 00:01.8-00:02.4: the dragon sits right beside the pot, inhales, its throat glows orange and its chest swells. 00:02.4-00:03.0: ONE short bright jet of fire, about 40 cm long, goes straight down into the open pot only. 00:03.0-00:03.5: the kernels burst at once into white popcorn that shoots out of the pot toward the lens and the woman's face. The firewood stays unlit, the wooden table does not catch fire. No dialogue in this block, only her sharp gasp. 00:03.5-00:06.5 SHOCK (one close-up, slightly handheld) Close-up of the woman's face at pot level: her whole face is covered in black soot except the area around her wide grey-blue eyes, popcorn pieces stuck in her hair and on top of her bun, a few popcorn pieces still falling past the lens. The baby dragon slides its head into the frame at her right shoulder, huge amber eyes, a guilty little smile, a thin puff of smoke from one nostril. She turns her head slowly toward the dragon and says in a stunned, breathless voice: "What did you do?" 00:06.5-00:09.5 THE SCOLDING (one medium-wide shot, three-quarter angle) The cottage door, the table, the wooden bowl and firewood are visible. 00:06.5-00:07.5: the woman straightens up and wipes her cheek with the back of her hand, smearing the soot. 00:07.5-00:08.5: she points one finger at the dragon and says firmly but with a hint of a smile: "Less fire." 00:08.5-00:09.5: she puts both hands on her hips. The dragon sits on the table among scattered popcorn and kernels, lowers its head, closes its eyes, folds its wings tight, the tail tip flicks nervously. At the end it opens ONE eye and peeks at her. 00:09.5-00:15 THE GENTLE FLAME (one medium shot, three-quarter angle) 00:09.5-00:11.0: the woman drops a quick handful of kernels from the wooden bowl into the pot, lifts it with both hands onto the iron tripod above the UNLIT firewood and closes the lid. The baby dragon sits on the table about 30 cm from the slab, on her right. 00:11.0-00:12.5: she raises ONE finger and says calmly, in a low, soothing voice: "Slow down... easy." while the dragon shuffles closer without leaving the table. 00:12.5-00:14.5: only now the dragon leans its head toward the firewood and breathes a thin, soft flame, careful and controlled. The kindling catches first, then the logs, and an orange ring of flame lights the base of the pot. This is the FIRST and ONLY fire under the pot, started by the dragon. The flame stays on the stone slab, the wooden planks do not burn. 00:14.5-00:15.0: the lid begins to rattle faintly as the first kernels pop inside. 00:15-00:18 PERFECT (one close-up, table level, held to the final frame) 00:15.0-00:16.2: low close-up behind the pot. Holding the folded linen cloth, the woman lifts the lid, steam rolls out and the pot is full of fluffy WHITE popcorn, none of it burnt. The dragon did NOT burn it. 00:16.2-00:16.6: her eyes go wide, her mouth opens in surprise. 00:16.6-00:17.7: she breaks into a bright happy smile, looks at the drag
Comece com um exemplo alinhado ao seu objetivo. Descreva claramente o tema e o resultado e acrescente composição, câmera, iluminação, áudio, movimento ou layout necessários para gerar Vídeo. Substitua as variáveis de produto, marca, referência e diálogo antes de gerar.
Abra um cartão, compare a prévia com o texto de origem e copie o Prompt original no Topview AI ou em um gerador compatível. Ajuste referências, restrições e notas de estilo ao seu projeto.
Use cada prompt como exemplo apresentado ou ponto de partida, não como garantia de resultado. Revise a precisão da marca, os direitos, a segurança e os recursos do modelo escolhido.
Sim. As páginas de mídia apontam para páginas de faceta específicas. Cada uma mantém uma única classificação ativa para preservar um escopo e uma intenção de busca claros.
A biblioteca inclui atualmente exemplos de Seedance 2.5, Seedance 2.0, Wan 3.0, GPT Image 2, Nano Banana Pro, Gemini Omni, Grok Imagine. A disponibilidade depende do tipo de mídia e dos filtros.
Você está vendo prompts de Vídeo da coleção “Person Character”, classificada por tema.