Kling AI 비디오 생성기

Replace the two men in the video with the two people from the reference images. Person 1 replaces the man on the left and Person 2 replaces the man on the right. Keep the original camera movement, timing, lighting, and body performance of both men, including their interactions and positions in the frame. Preserve Person 1's bald head, beard, black glasses, white t-shirt, light blue jeans, white sneakers with burgundy detail, and gold chain. Preserve Person 2's face, hair, and outfit exactly as shown in the reference images. Keep both identities distinct and consistent throughout the video, and do not mix or swap their features.

Scene & Mood: Cinematic luxury portrait in a warm amber Arabian lounge. Calm, powerful, mysterious mood. Rich red textiles, brass lanterns, patterned cushions, soft haze, and practical firelight. Frame Map: Start from image1composition. Subject stays center right in a medium wide seated portrait, facing camera. Knees and hands remain visible. Ornate tea table and brass objects stay softly blurred in the lower foreground. Lanterns, cushions, and carved wall decor remain behind him. Keep warm negative space on the left. Slow push in ends as a closer medium portrait. Subject Lock: Image1 Preserve the same seated male subject: same face, skin tone, trimmed facial hair, dark sunglasses, red and white keffiyeh, black agal, white thobe, leopard pattern cloak, watch, rings, body shape, seated posture, relaxed hand placement, and composed expression. Image 1 the same face, hair, wardrobe, body shape, and silhouette throughout. Cloak stays naturally draped over both shoulders. Thobe stays matte white with soft real fabric folds. Cross Frame Rules: No identity drift, no face change, no wardrobe change, no altered sunglasses, no extra jewelry, no headwear change. Subject remains center right at the same depth and eyeline. Do not add people or modern objects. Keep the same lounge layout, lantern positions, patterned textiles, brass objects, amber palette, and shallow depth. Movement: 15s continuous shot. Subject remains seated and still with slow breathing. His right fingers relax slightly on his knee, left hand barely shifts against the fabric, and his head tilts a few degrees toward camera. Keffiyeh edge, cloak fibers, thobe folds, watch, and rings catch tiny natural movements. Lantern flames flicker gently, brass highlights shimmer softly, haze drifts in the background. Camera makes a slow, smooth push in with subtle handheld breath, never shaky. Last Frame: End on a closer medium portrait. Subject remains center-right, sunglasses facing camera, shoulders and upper torso dominant, right hand relaxed on knee, leopard cloak framing the white thobe. Lantern glow and patterned textiles stay softly visible behind him. Camera settles nearly locked. No on screen text, no captions, no signage typography, no rendered text in the frame. World Plate: Ornate lounge with burgundy, red, black, gold, and amber tones. Layered rugs, cushions, wall hangings, carved circular wall decor, brass lanterns, and polished metal objects create depth. Lighting comes from practical lanterns with soft falloff and lifted shadows. Mild incense haze adds air depth. Textures feel tactile: woven fabric, matte cotton, aged brass, carved metal, natural looking fur. Sound Bed: Diegetic audio only: quiet lantern flame flicker, faint room tone, soft fabric rustle from breathing, subtle jewelry movement, distant air movement. No music, no score, no vocals, no lyrics. Capture Realism: Natural skin texture, matte skin, realistic facial hair, believable fabric weight, soft highlight rolloff, lifted shadows, mild grain, restrained contrast, no plastic AI gloss, no doll skin, no over sharpening, no crushed blacks, no clipped whites. Brass glows softly, not mirror shiny. Fur texture stays natural. Camera Capture: 50mm cinema lens, slow push in portrait shot, 24fps, 180 degree shutter, wide latitude cinema capture, subtle handheld breath, shallow depth of field, fine 35mm film grain, warm amber burgundy grade, realistic practical lantern light, 15s runtime.

{ "description": "A high-speed cinematic FPV drone shot begins far above dramatic desert cliffs at early dawn. The drone dives steeply into rugged rocky canyons, skimming over sand dunes and racing between massive red stone walls. Wind lifts trails of dust as the drone accelerates, narrowly passing through natural rock arches and sunlit ridgelines. Midway, the motion gradually slows as the drone bursts out onto a wide open desert plateau, revealing a single ice-cold Red Bull can standing on a rugged stone slab, condensation visible and sparkling in the morning light. No text.", "style": "cinematic, hyper-realistic extreme sports commercial", "camera": "first-person FPV drone perspective with aggressive high-speed canyon flythrough, transitioning into a smooth, controlled slow cinematic reveal", "lighting": "early sunrise light with warm highlights, strong rim lighting on rock formations, high contrast shadows; crisp reflections and condensation highlights on the Red Bull can", "environment": "vast desert canyon landscape featuring red rock cliffs, narrow canyons, sand dunes, wind-blown dust, and an open plateau clearing", "motion": "very fast aerial movement with sharp dives, close canyon passes, and tight turns; gradually transitions into slow motion and smooth stabilization for the final reveal", "ending": "a powerful, quiet moment on the open desert plateau; the Red Bull can centered on the stone slab as the camera gently orbits and pushes in, emphasizing energy, freshness, and impact", "aspect_ratio": "16:9", "duration": "10 seconds", "text": "none", "keywords": [ "FPV drone", "ultra fast cinematic motion", "desert canyon flythrough", "hyper-realistic product reveal", "energy drink commercial", "slow motion ending", "no text" ] }

Create a seamless 15-second cinematic fantasy action sequence using @ Image 1, @ Image 2, and @ Image 3 as consecutive visual keyframes. Preserve the same moss-covered stone colossus, forest geography, riders, atmospheric lighting, ancient architecture, scale, and golden energy throughout. 0 to 5 seconds, @ Image 1: Begin instantly with an epic ultra-wide aerial view of the colossal forest valley. The camera rushes forward just above the glowing meadow as four riders gallop toward the advancing moss-covered colossus. Massive footsteps shake the ground, leaves swirl through the air, distant waterfalls flow, lightning flashes behind the trees, and clouds of mist rise around its legs. Maintain immense scale and realistic horse movement. 5 to 10 seconds, @ Image 2: Sweep rapidly between two gigantic tree trunks and transition into a dramatic low-angle tracking shot. Follow the riders and lone figure racing beneath the colossus as one enormous foot strikes the forest floor. Moss-covered stone fractures, roots split, boulders tumble, branches fall, and green dust erupts toward the camera. Introduce brief slow motion as debris, spores, leaves, and fragments float around the characters, then accelerate forward through the opening beneath the colossus. 10 to 15 seconds, @ Image 3: Snap back to full speed and surge upward along the colossus body. Break through clouds and waterfalls to reveal that its enormous head contains a hidden golden city. Orbit around the floating stone face as illuminated towers, bridges, temples, windows, and waterfalls emerge from the cracked skull. End with the camera diving through one glowing opening into the city while golden light fills the frame. Use photorealistic cinematic detail, grand fantasy scale, smooth spatial continuity, dynamic parallax, volumetric forest light, realistic environmental destruction, natural motion blur, controlled camera shake, and detailed water simulation. Keep the colossus design consistent. No character duplication, sudden costume changes, deformed horses, unstable architecture, random scene changes, flickering details, text, logos, or desert imagery.

A giant, colossal, and aggressive creature in a turbulent ocean amidst massive waves and a storm.

INFINITE FALL — Beer Catch [image1] Guy: lean young man, wavy blonde hair, thin glasses, black short-sleeve open shirt, dark grey pinstripe wide-leg trousers, large silver belt buckle, brown leather loafers, arm tattoos, silver bracelets. Expression: not panicked — annoyed at the bottle, nothing more. IMAX 65mm aggressive handheld — camera itself in freefall with him, shaking slightly with air resistance, never still. Fast orbital cuts, whip pans between angles, camera sometimes rushing past him at speed. Everything blown violently upward: hair, jacket, trouser legs, shirt — continuous strong upward wind throughout. The visual grammar of terminal velocity. Sound FX: constant roaring wind, fabric snapping, bottle tumbling, liquid sloshing. Environment: pure white infinite void — no ground, no ceiling, no horizon. Just white in all directions. A brown glass beer bottle tumbling slowly in freefall beside him, liquid sloshing inside, just out of reach. negative: no ground, no slow camera, no calm floating — this is a violent fall The entire sequence is one continuous fall at full speed. Camera whips around him — bug's eye from directly below looking up at him shrinking then rushing toward camera, overhead looking straight down at him, 10mm close pass inches from his face with the wind tearing at his hair and glasses, Dutch angle tracking his full body as he reaches. The bottle tumbles just past his fingers — he repositions fast, body rotating in the fall, reaching again. His glasses nearly torn off by the wind, he pushes them back with one hand without stopping reaching with the other. The fall feels violent and fast — not peaceful zero gravity, not serene floating. The roaring wind is constant and loud. His clothes are being destroyed by the air resistance — shirt billowing completely open, trouser legs flattened against his legs then snapping outward. He finally grabs the bottle in one hard grip — fingers closing around the neck. He pulls it against the wind to his lips. Takes a long defiant sip. Wind still roaring. Hair still destroyed. Glasses still crooked. Camera does one final fast orbit around him — close, aggressive, whipping past. He lowers the bottle. Looks directly into the camera lens from centimeters away. Completely unbothered. Wind roar continues. Cut to black. IMAX/anamorphic, violent freefall camera, aggressive orbital handheld, constant roaring wind, infinite white void, 4K Music: Flex It - MPH, y3kfae

@ Image1 is the lead performer — preserve her exact identity in every shot: cornrow braids, septum ring, statement earrings, sculptural glossy white wrapped top, dark indigo denim. @ Audio1 is the music track — entire edit cut to it; no invented music, no voiceover, no lip-sync. Music-video route: dance/performance, brutalist stage show. Track: industrial hip-hop, 95 BPM half-time, distorted 808s, luxury-noir. Director thesis: architecture as choreography — dancers staged on concrete plinths at different heights, a human sculpture garden that activates around the lead. Visual world: brutalist concrete atrium with massive raw concrete plinths and steps at staggered heights, shafts of cold daylight from a skylight, dust motes in the light shafts, dark shadow pockets. Background dancers: 6 dancers in oatmeal and bone-toned knitwear and tailoring, sculptural silhouettes, like museum figures. Palette: bone, oatmeal, concrete grey, deep indigo accent on the lead. No neon, no particles beyond natural dust. Cut map synced to @ Audio1: 0.0–2.0s THE BREATH — static ultra-wide symmetrical: six dancers frozen in sculptural poses on six plinths at different heights, the lead standing on the floor at center like a visitor. Total stillness, museum silence. 2.0s THE DROP — 808 kick: every statue snaps into a new pose simultaneously on the hit, heads turning to her; hard cut to a fast push-in on the lead. 2.0–4.0s she struts between the plinths, camera tracking laterally through foreground columns creating parallax wipes; on each kick one statue above her snaps a new angular pose. 4.0s snare — whip-pan: full freeze, she hits a profile vogue arm frame between two plinths, statues frozen mid-rotation above. 4.0–8.0s sculpture canon: her sharp head isolations and New Way vogue lines on the snares trigger cascading pose-changes across the plinths, a mechanical wave of bodies jumping pose to pose like stop-motion statues; torso rolls on the 808 slides travel up the plinth heights from low to high. Camera slowly cranes upward through the atrium. 8.0s hi-hat triple roll — 3 rapid macro inserts: her earring in a light shaft, a dancer's hand rotating on a plinth edge, dust exploding off concrete as a boot lands. 8.5–12.0s escalation: accelerating orbital-crane move spiraling around the whole sculpture garden as all six dancers now move continuously in their own orbits on the plinths, her arm work at the center speeding up; speed-ramp stretch then snap back on the kick. 12.0s bass slide — the dancers step down off the plinths in unison and sink into deep lunges around her as she glides through a sideways level drop, camera dutch and low across the floor. 13.5–15.0s FINAL BUTTON — hard cut to the opening ultra-wide: she now stands ON the tallest central plinth in a frozen vogue pose one arm high, the six dancers frozen as statues on the floor below looking up; micro push-in, hold. Loopable. Performance rules: lead stoic, dominant; dancers precise, sculptural, expressionless. Continuity: same seven people, same wardrobe, same atrium. Audio intent: every cut and pose-snap on a hit of @ Audio1; faint foley only (steps, fabric, room tone). Quality bar: expensive editorial music video, museum-grade art direction, no AI gloss, no glow.

**CAMERA:** DV 16mm handheld camcorder POV. CHASE holds the camera or props it on the treadmill/bike console. Natural shake, soft focus, delayed focus pulls, clumsy zooms, imperfect framing. Camcorder never visible. **LOOK:** Soft vintage tape quality, slight blur, tape noise, bloomed gym lights, flickering auto-exposure, muted contrast. **STYLE:** High-energy cardio vlog. Upbeat, breathless, camera bouncing with movement, building to a sweaty, satisfying finish. **CHARACTER:** CHASE — Korean woman in her 20s. High ponytail, glowing skin with increasing sweat, slim athletic build. Modest long-sleeve athletic top, loose joggers/leggings, sneakers. **SETTING:** Evening gym cardio area with treadmill/bike, mirrors, cardio machines, phone playing music, soft lighting. **SCENES:** 1. Starts treadmill. "Okay, cardio time!" 2. Running. "This song's so good." 3. Wipes sweat. "Five minutes in and I already regret this." 4. Close-up of feet, sweat, treadmill. 5. Faster sprint. "Almost... done..." 6. Slows down, towel. "I survived." 7. Sweaty grin. "That actually felt really good." 8. Steps off, thumbs up. "Cardio done—see you guys."

Photorealistic full-body reference portrait of a beautiful young woman snowboarder, standing on a snowy mountain slope, bright daylight, sharp cinematic detail, 8k resolution. She has pale skin, striking ice-blue eyes, long straight black hair flowing down past her shoulders, delicate feminine facial features, high cheekbones, natural makeup. She looks directly at the camera with a calm, confident, fearless expression. She is wearing a complete feminine professional snowboarding outfit: • Fitted insulated snow jacket in matte black with subtle teal accents and white geometric patterns • Matching black snow pants with reinforced knees • Black helmet with teal details • Goggles resting on top of the helmet • Black snow gloves • Black snowboard boots She is standing on a modern freestyle snowboard: black base with teal and white graphic design, visible bindings, slightly angled toward the camera so both the board and her full outfit are clearly visible. Natural standing pose, one hand on her hip, relaxed athletic posture, weight slightly on the back foot. Environment: fresh white powder snow around her, dramatic alpine mountain peaks in the soft background, bright sunlight with subtle lens flare, crisp cold atmosphere, photorealistic snow texture, ultra-detailed fabric and material textures, realistic hair strands, sharp focus on face and outfit. Style: hyper-realistic photography, shot on Sony A1 with 85mm lens, f/1.8, natural lighting, National Geographic level detail, no text, no logos, no watermarks. Negative prompt (if supported): blurry, deformed, extra limbs, low quality, cartoon, illustration, painting, text, watermark, overexposed, underexposed, plastic skin.

FPV shot: The camera dives through a tropical waterfall hidden inside a lush jungle. Water crashes through the frame while exotic flowers, mist, and sunlight fill the canyon. The camera races through the spray as billions of droplets rise into the air and naturally form "PARADISE" before dissolving into a radiant rainbow. Exotic, luxurious, breathtaking nature.

Create a 15 second photorealistic live action comedy in a horizontal 16:9 frame, 4K detail, 30 frames per second. The lead is a strikingly attractive adult East Asian woman in her late twenties with long glossy black hair in loose waves, elegant natural makeup, expressive dark eyes, and a confident sensual screen presence. She wears a fitted midnight blue satin lounge dress with a tasteful neckline and turquoise socks. She holds an ornate navy and gold hardcover book respectfully against her chest. The cover has decorative geometry only, with no readable writing. The location is a real lived in townhouse bedroom connected to a narrow hallway. Show warm honey oak floors, cream walls, framed art, a beige patterned bed, a grey velvet bench, a faded Persian rug, sheer curtains glowing with daylight, a cast iron radiator, and a vintage brass and white ceiling fan. Place one heavy black medicine ball in the hallway. Keep every object, facial feature, garment, and room detail consistent through all cuts. From 0 to 3 seconds, use a locked wide camera at waist height with a natural 24 millimeter phone lens look. She walks toward camera while calmly reading, centered in the hallway. Her socked foot clips the medicine ball. She stumbles with believable weight and momentum, instinctively protects the book, and falls sideways onto the bed. Show realistic hair inertia, fabric folds, foot placement, motion blur, and mattress compression. From 3 to 6 seconds, cut to a close side angle near the bed. She sits upright without losing her place, still reading with total focus. The medicine ball rebounds through the room. She leans back just before it passes, then calmly pulls the blanket over herself while keeping the book open. Use quick hard cuts and one brief imperfect whip pan timed to the impacts. From 6 to 10 seconds, cut to a low upward angle of the slowly rotating ceiling fan. The ball arcs into the blade path and is redirected through a chain of physically believable collisions. The fan wobbles subtly, dust shakes loose, and each impact has accurate contact, spin, gravity, and rebound. Cut to the radiator as the ball strikes the metal ribs, drops to the wooden floor, and rolls out of frame. From 10 to 15 seconds, return to a wider bedside composition. She lies comfortably under the blanket, still reading and completely unbothered. The ball rolls back into the foreground, lightly bumps the bed frame, and stops. She gives the camera one brief deadpan glance, then returns to the page. Hold the final image for one second. Use soft window daylight, gentle bounced fill, realistic shadow falloff, small phone auto exposure shifts, natural skin texture, slight lens distortion, deep focus, and restrained handheld vibration. Avoid polished commercial perfection. Edit with fast impact cuts while preserving screen direction. Add playful music around 129 beats per minute with dry percussion, plucked bass, light handclaps, and a mischievous rising rhythm. Layer rubber ball thumps, a bed creak, fabric movement, fan clacks, radiator clangs, footsteps, and quiet room tone. Briefly lower the music at every impact. No dialogue, no captions, no logos, no watermark, no visual glitches, no morphing, no extra fingers, no duplicated objects, and no impossible body movement.

**SCENE** Seedance 2.0 slice-of-life, 15 seconds. @ image1 = Main Subject. Preserve exact face, hairstyle, features, skin tone, and body proportions throughout. **OUTFIT:** Oversized olive sleeveless linen shirt loosely tucked into wide-leg dusty burgundy trousers, flat brown leather sandals, small gold hoop earrings. Shirt naturally damp/sweat-stained at collar, underarms, chest. Hair loosely down, natural waves, few strands sticking to face and neck. Completely unglamorous. **LOCATION:** Authentic traditional Korean rural village. Dirt/gravel paths. Mossy stone walls. Hanok houses with aged clay tile roofs. Wooden fences. Overgrown grass. Gnarled old trees. Stone water basin. Hanging dried chili peppers. Cracked stone steps. Ceramic jangdok jars. Free-roaming roosters. Morning cooking smoke. Dense hillside vegetation. Zero modern elements. **CAMERA:** Late-2000s flip cam vlog. Heavy handheld shake. Imperfect reframing. Subject frequently off-center or cropped. Aggressive autofocus hunting. Lens breathing. Strong exposure swings. Rolling shutter. Heavy motion blur. Warm faded washed-out muddy tones. Blown highlights. Crushed blacks. Heavy digital noise and compression artifacts. Maximum home-video imperfection. --- **00:00–00:02** — Mossy stone wall path. Loose sweaty arm's-length self-shot. Woman smiles tiredly, speaks casually to lens while walking. Weeds brush past. Camera tilts mid-walk. **00:02–00:04** — Hanok courtyard. Handheld from behind. Woman walks through; stray dog trots alongside. Jangdok jars along wall. Morning smoke drifts low. **00:04–00:06** — Stone water basin. Handheld close. Woman crouches, splashes cool water on sweaty face and neck, exhales with relief, laughs softly. Focus hunts aggressively. **00:06–00:08** — Gnarled tree, stone wall. Low handheld. Woman crouches, hand-feeds stray dog. Camera breathes between her face and the dog. **00:08–00:10** — Village path. Handheld side tracking. Woman walks slowly, wipes sweat from forehead, glances around. Rooster crosses ahead. Camera sways, reframes late. **00:10–00:12** — Shaded hanok exterior, stacked ceramic pots. Woman sits on cracked stone step, sips from battered tin cup. Shirt visibly damp. Elderly woman moves in background. Breeze lifts hair. **00:12–00:15** — Dirt path near gnarled tree. Self-shot. Woman raises camera to sweaty face, laughs tiredly, speaks casually, gives small wave. Recording cuts abruptly mid-moment. --- No posing. No fashion look. No perfect composition. No stabilization. No cinematic glamour. No modern grading. No music. **AUDIO:** Morning birds, roosters, distant village chatter, dog, water on stone, crackling fire, footsteps on dirt, rustling leaves, wind, ceramic sounds. **16:9.**