Seedance 2.5 AI 视频生成器

Create a product video ad from @image1,@image2, script "A high-end cinematic fashion ad video, vertical 9:16, ultra-realistic, premium lighting, smooth transitions, strong rhythm. Use the exact female model shown in the reference as the model and the exact outfit shown in the reference as the outfit, worn consistently in all shots with full design, color, material, and structure preserved. Maintain consistent identity. Industrial city aesthetic, cool tones shifting to warm sunset. No text, no watermark. 0–2s (Hook + Detail) Indoor minimalist space. The exact female model shown in the reference walks away from camera, soft natural light. Cut to close-up of outfit fabric, folds, and structure, realistic texture. 2–4s (Lifestyle) Wide shot in industrial corridor, full outfit visible. Side tracking shot along concrete wall, emphasizing silhouette and structure lines. 4–6s (Detail + Portrait) Macro of fabric movement (layered/flowing). Cut to mid shot portrait, calm confident expression, slight hair movement. 6–8s (Detail + Lifestyle) Extreme close-up of stitching, seams, material transitions. Top-down wide shot walking through dark industrial tunnel, strong contrast. 8–10s (Detail + Lifestyle) Tracking shot of garment motion while walking. Transition to warm sunlight street crossing, enhancing texture and color. 10–12s (Detail + Portrait) Close-up of lower body or hem movement. Cut to face close-up, clear skin detail, steady gaze. 12–15s (Final) Epic wide sunset shot on a bridge. The exact female model shown in the reference faces skyline, then looks toward camera. Warm light, slow-motion ending. Style: cinematic fashion ad, high-contrast lighting, realistic motion, sharp details, premium commercial quality, no distortion."

Use the uploaded reference image as the exact character reference. Preserve her facial identity, grey-green eyes, skin tone, hairstyle, makeup, body proportions, and overall appearance throughout every shot. She has long dark hair in a high ponytail and wears a fitted black long-sleeve athletic crop top, high-waisted black leggings, clean white sneakers, and a white towel around her neck. Maintain perfect character consistency. Create a realistic late-night fitness vlog filmed on a handheld smartphone with a subtle vintage DV/camcorder aesthetic. The gym is quiet and nearly empty, with large mirrors, dumbbell racks, benches, stretching mats, warm overhead lighting, and realistic reflections. The video begins with an arm's-length selfie shot as she walks into the gym, slightly breathless but smiling at the camera. She casually pans across the empty gym before returning to her face. She stops near a mirror, fixes her ponytail, gives herself a playful look, then picks up dumbbells. Cut to a fixed external angle as she performs controlled shoulder presses with realistic form and natural breathing. Return to handheld selfie footage as she walks with the towel around her neck, looking tired but satisfied. She grabs a water bottle, takes a drink, and smiles at the camera. Cut to a wider fixed shot where she performs a quick stretch and playful mirror pose. Finish with a close selfie as she brushes back a loose strand of hair, smiles, waves toward the camera, and walks toward the gym exit. Camera: realistic handheld shake, slight framing imperfections, delayed autofocus, occasional motion blur, subtle exposure shifts, natural jump cuts, no visible camera or camera setup. Style: authentic creator vlog, glamorous fitness influencer aesthetic, realistic body movement, natural expressions, soft DV texture, subtle tape noise, warm cinematic lighting, shallow depth of field, photorealistic, 4K HDR, 16:9. Audio: natural gym ambience, footsteps, dumbbells, breathing, water bottle sounds, soft room tone. No background music, no subtitles, no captions, no logos, no watermarks. Negative Prompt: No face changes, identity drift, outfit changes, extra people, duplicate limbs, distorted hands, broken gym equipment, impossible movements, artificial skin, excessive beauty filters, flickering, cartoon style, visible camera.

30s | 16:9 | Realistic UGC cleanser recommendation | strict one-take | natural smartphone footage Use the same young woman from the reference image throughout the entire video. She has a short black bob haircut, light natural makeup, real skin texture, a slim body, and natural proportions. Keep her face, hairstyle, outfit, and body proportions fully consistent in every second of the video. She must wear the same blue short-sleeve collared pajama top and matching blue pajama shorts from the reference image. The product is one facial cleanser. The key product behavior is: when mixed with water in her hands, it should create rich, fine, realistic white foam. The foam must be formed naturally through rubbing with wet hands, then applied onto the face. It must not appear as a visual effect or pop into existence. Scene A cozy, ordinary apartment with natural daylight. The video starts in the living room. Background includes a beige sofa, coffee table, floor lamp, green plants, and a few real everyday items. The girl sits cross-legged on a rug. In front of her is a plain kraft shipping box, and beside it is a small household pair of scissors. The bathroom is connected nearby and can be naturally walked into. Spatial continuity must feel real and uninterrupted. Camera style This must be one continuous take from start to finish. No cuts, no transitions, no hidden edits, no jump frames, no teleportation. It should feel like a friend is filming her on a phone. Use a natural smartphone wide-angle lens, with slight handheld shake, realistic autofocus and exposure changes. Do not make it feel like a polished commercial. No cinematic camera moves, no subtitles, no over-retouching, no plastic skin. The overall vibe: she just received a cleanser her friend recommended and casually filmed a real unboxing + first try. Sequence 0–3s | Opening Camera starts in the living room. She sits on the rug with the package in front of her, looks at camera, smiles naturally and says: “Okay, this finally came.” Then she looks down at the package. 3–8s | Unboxing She picks up the scissors, holds the box steady with one hand, and cuts through the top tape naturally while saying: “My friend has been telling me to try this cleanser.” She puts the scissors back down and opens the box with both hands. 8–12s | Product out She takes the cleanser out, looks at it briefly, then lifts it naturally toward the camera. No cutaway product close-up, only a slight camera lean-in. She says: “She literally swears by it.” Then smiles and says: “Let’s try it.” She prepares to stand up with the cleanser. 12–17s | Walk to bathroom She stands up naturally while holding the cleanser. The camera backs up and follows her continuously as she walks from the living room into the bathroom. She says while walking: “I haven’t even washed my face yet, so perfect timing.” 17–20s | Wet face At the sink, she places the cleanser beside the sink, turns on the tap, gathers water in her hands, and wets her face. Water must physically touch the skin. Natural droplets appear on hands and face. 20–23s | Foam up She picks up the cleanser, squeezes a small amount into her wet palm, puts the cleanser back down, adds a little water, and rubs both hands together. The foam must build gradually from a small amount into rich, fine white foam. 23–26s | Wash face She applies the foam to her face and massages her cheeks, nose, forehead, and chin naturally. Foam follows her hand movement realistically. Do not cover her eyes. Do not make the foam excessive or like shaving cream. She reacts with mild pleasant surprise and says: “Oh... this feels really nice.” 26–28s | Rinse off She leans toward the sink, uses clean water to rinse off the foam. The foam gradually disappears with water. Her face is revealed clean and naturally wet, with a few droplets still visible. 28–30s | Final reaction She lifts her head, gently touches her cheek, turns to the camera, smiles with a pleasantly surprised expression and says: “Okay... I get the hype.” Then she picks up the cleanser and shows it naturally, saying: “She was right.” End naturally. Critical continuity rules The whole action must happen in one real continuous shot: sitting in living room → pick up scissors → cut tape → put scissors down → open box → take out cleanser → show cleanser → stand up → walk to bathroom → place cleanser down → turn on water → wet face → squeeze cleanser into palm → add water → rub hands → foam forms → apply foam to face → massage → rinse off → touch face → pick up cleanser → share reaction. Must avoid No second cleanser. No product teleportation. No hairstyle change. No outfit change. No face change. No hand/finger deformities. No environment changes. No sudden dry/wet face continuity errors. No foam appearing from nowhere. No hidden cuts. Key story: friend strongly recommends it → package just arrived → she opens it immediately → walks straight to the bathroom → tries it right away → real foaming wash → genuinely likes how her skin feels. Natural, casual, relaxed, everyday American UGC tone.

A young woman stands alone on a rainy Seoul street at night, holding her smartphone as neon signs reflect across the wet pavement. She notices a missed call from an unknown number, then answers when the phone rings again, slowly becoming uneasy as silence fills the call. She turns and sees a young man under a black umbrella staring at her with a quiet smile, but a passing car briefly blocks her view, and when it clears, he has mysteriously disappeared. Her phone lights up with a new message: “Don’t look back.” The scene ends on her frightened expression, with realistic Korean actors, cinematic rain, moody blue and warm lighting, shallow depth of field, subtle film grain, natural acting, suspenseful pacing, and premium Korean thriller cinematography.

Use the uploaded storyboard image as the exact visual reference for the sequence. Use the uploaded character image as the exact character reference. Preserve her facial identity, eye color, skin tone, long straight black hair, body proportions, makeup, and outfit consistently throughout every shot. She wears a hot-pink leather biker jacket over a white crop top, layered silver chain necklaces, gold star earrings, deep red lipstick, and dark oversized pants. Maintain the same recognizable performer in every scene. Create a 15-second ultra-realistic Y2K rap/pop music video inspired by early-2000s music television aesthetics, filmed with an exaggerated ultra-wide fisheye lens. The visuals should feel energetic, glossy, slightly surreal, and heavily stylized, with dramatic lens distortion, dynamic camera movement, and fast beat-synchronized editing. — Light-Up Reveal: Begin in complete darkness inside a long futuristic black corridor. The rectangular wall lights suddenly switch on one after another, creating a rapid chase effect toward the performer. She stands completely still in silhouette with her arms spread. On the final light hit, the overhead lights illuminate her face as she immediately begins lip-syncing directly toward the camera. — Fisheye Close-Up: Hard cut to an extreme fisheye close-up of her face. She leans aggressively toward the lens, performs confident hand gestures, points directly at the camera, and briefly pushes her palm toward the lens, making her hand appear dramatically oversized from the fisheye distortion. Chrome Room: Cut to a glossy futuristic white-and-chrome room. She performs a playful, confident dance with shoulder movements, small footwork, and sharp hand gestures. The camera smoothly slides sideways at knee height while chrome surfaces create strong reflections and parallax. 6–8s — Graffiti/Dollar Set: Hard cut to an abstract orange-and-lime studio filled with oversized glossy black geometric shapes featuring generic dollar symbols. She performs in profile while lip-syncing, then suddenly turns her head toward the camera on the beat. Her ponytail and jacket move naturally with the turn. 8–10s — Dancer Performance: Reveal two anonymous female backup dancers wearing glossy black full-coverage futuristic outfits and dark visor sunglasses. They flank the performer and perform synchronized Y2K choreography while she dominates the center, pointing toward the lens and bouncing confidently to the rhythm. 10–12s — Beam Dome: Cut to a dramatic low-angle shot. She stands on a small pedestal inside a dark futuristic dome as powerful white light beams switch on around her one at a time. She raises both arms and continues performing while the lights create intense silhouettes and lens bloom. — Chrome Macro: Extreme tilted fisheye close-up of half her face against a quilted metallic-silver wall. She gives a playful smirk and delivers a

Replace the main character in the white model video with an image to generate a video of the woman in black standing on a cliff in a vast valley. The scene starts with a front shot of the female lead, with an orbiting camera as she picks up a flask to drink. As the camera orbits to her back, a massive white dragon floats in the sky. Realistic texture, cinematic feel, generate sound effects only, no background music.

The Concept: Jetpack Chameleon's Toybox Heist Visual Style: High-end 3D cartoon animation (Pixar/Spider-Verse hybrid style), vibrant neon and brass colors, volumetric lighting, motion blur for speed, and hyper-dynamic camera angles. The Animation Prompt (15-Second Breakdown) [0:00 - 0:03] The Explosive Escape Action: No hesitation. The scene opens mid-action. A bright turquoise chameleon wearing a clunky, spark-spitting steampunk jetpack bursts upward out of a giant, glowing jack-in-the-box. He clutches a glowing, levitating crystal key in his tail. Camera: The camera is positioned low, pointing up, then instantly snaps back to track him from over-the-shoulder as he rockets forward, leaving a trail of purple smoke. [0:03 - 0:07] The Obstacle Course Action: The chameleon dives into a chaotic, towering cityscape made of stacked board games and tumbling dominoes. He barrel-rolls through a looping, neon-orange racetrack, narrowly dodging a barrage of glowing bouncy balls that crash around him like meteors. Camera: Fast-paced tracking shot. The camera corkscrews along with the chameleon, maintaining high speed. A sudden "whip-pan" to the right reveals the threat. [0:07 - 0:11] The Pursuer Strikes Action: A massive, robotic origami dragon with glowing red LED eyes bursts through a tower of playing cards right behind him. It snaps its jagged paper jaws, barely missing the chameleon’s tail. The chameleon looks back, grins, and hits a red button on his jetpack. Camera: Extreme close-up on the chameleon's wide, crazy eye, reflecting the glowing red dragon. A rapid zoom out as his jetpack violently backfires a burst of confetti directly into the dragon's face. [0:11 - 0:15] The Slingshot Finale Action: Blinded, the dragon crashes. The chameleon shoots a grappling hook from his wrist onto a spinning desk globe. He uses the momentum to swing a full 360 degrees around the globe, slingshotting himself forward at warp speed. He shatters through a giant stained-glass window, exploding out into a vibrant, sunlit sky amidst a shower of colorful, glittering glass shards. Camera: The camera follows him around the globe, then stops dead as he launches toward the window. Slow-motion for exactly one second as the glass shatters, then snaps back to real-time as he vanishes into the clouds.

A 15-second cinematic sci-fi epic movie-trailer montage on a vast desert planet, anamorphic widescreen 21:9, ultra-cinematic large-format look, vintage anamorphic lenses, shallow depth of field, thick atmospheric dust haze, strong horizontal lens flares, warm sand-amber highlights against deep teal shadows, filmic grade, fine 35mm grain, monumental awe-inspiring scale, no on-screen text, no watermark. A sequence of hard cuts like a prestige film trailer, accelerating toward a rapid-fire climax. SHOT 1 (0.0-2.0s): Near-black frame of slow-drifting golden dust and dark teal atmosphere, a single soft horizontal anamorphic light streak, faint glowing horizon far below, embers rising. Extremely slow push-in. Then it SLOWLY DISSOLVES (not a cut) into: SHOT 2 (2.0-4.5s): Tight low-angle portrait of a bald, weathered desert leader in coarse pale robes, sun-cracked skin, calm powerful gaze; behind him, out of focus, a monumental sand-colored ruined desert city dissolving in haze. Very slow dolly-in, warm rim light, drifting particles. SHOT 3 (4.5-5.6s): HARD CUT to a colossal circular arched portal hall, perfectly symmetrical, a long central causeway to a giant glowing round gate, rows of tiny robed figures lining both sides, warm god-rays and dust. Static wide, overwhelming scale. SHOT 4 (5.6-6.6s): HARD CUT to a dark-haired bearded man in a pale robe on a pure black background, one hard side light, he slowly bows his head, dust motes in the beam. Chiaroscuro, intimate. SHOT 5 (6.6-7.5s): HARD CUT to a single lone hooded robed figure standing in an almost pitch-black void, one faint edge of cold light on the robe, drifting dust. Minimal, ominous, still. SHOT 6 (7.5-8.6s): HARD CUT to a vast cathedral-like hall, hundreds of robed figures in two long symmetrical rows flanking a white central aisle, a small group of hooded figures walking slowly toward camera down the middle, towering angular architecture, god-rays and haze. Slow dolly. SHOT 7 (8.6-9.5s): HARD CUT to a woman in flowing white robes walking across open desert toward a sleek curved chrome-white aircraft on the sand, crowds of robed people, blinding sun flare off the hull. Tracking from behind, blowing sand. SHOT 8 (9.5-10.85s): HARD CUT to an extreme close-up of the heroine's face wrapped in a dark desert scarf, only her eyes and cheekbones lit, an intense determined stare into the lens, a bright circular halo of backlight glowing behind her head, soft anamorphic bokeh. Barely-moving push-in. Emotional center. SHOT 9 (10.85-12.4s): HARD CUT to a dark elongated bird-like aircraft skimming fast and low over rolling dunes at golden dusk, silhouetted against the sun, sand kicked up in its wake. RAPID CLIMAX MONTAGE (12.4-15.0s) — a burst of very short hard cuts, roughly 0.3s each, escalating impact: 12.4s: wide smoking battlefield across the desert, a crashed burning airship, soldier silhouettes, thick smoke. 13.1s: same battlefield closer, a dark huma

テンポが良く、毎秒60フレームの高フレームレートによる高品質な動画を生成してください\n\nショットとアクションの演出：\n\n0-2s：廊下のワイド、 📷画像2 （トラ）と 📷画像3 （マル）が同時に叫ぶ出だしで身体が声の頭で前へ伸び、尻が一拍遅れて追いつく、カメラは低い位置で前方へ急加速して牽引。\n\n2-6s：廊下をジグザグに爆走、足は円形の回転残像になり、背後に紙と綿の長い尾が引きずられる、カメラは左右に切り返しながら追走。\n\n6-9s：曲がり角の中景、 📷画像3 （マル）だけ壁へ平たく押し潰される、犬の形が一瞬ペラいシルエットになるのを正面から見せる。\n\n9-10s：同構図でポンと元へ戻る瞬間を一拍で抜き、すぐ走りへ復帰。\n\n10-14s：床上すれすれの追跡、 📷画像2 （トラ）がベッド下へ液体のように薄くなって滑り込む、カメラも高さを落として隙間へ突入。\n\n14-18s： 📷画像3 （マル）が頭から突入するが腰で止まり、身体がアコーディオン状に縮む、カメラは横から寄って伸縮を誇張。\n\n18-21s：一度反動で飛び出し、もう一度突っ込む、同じ動きをテンポよく繰り返し、最後は上半身だけが入って尻と巻き尾が外に残る。\n\n21-24s：ベッド下の暗闇の主観寄り、暗闇から 📷画像2 （トラ）の前足が伸びて 📷画像3 （マル）の頭を必死に引く、毛が引っ張られても1ミリも動かない。\n\n24-27s： 📷画像3 （マル）の尻側の中景、外に残った尻と尾が小刻みに震え、床の綿と紙が擦れて動く。\n\n27-30s：暗闇の奥から二匹の目だけが見える寄り、目が焦点を失ったまま固まり、引っ張りの力が抜けて“詰み”の間でカット。\n\nキャラクターは storyboardDialogue に記載された台詞を一字一句そのまま話し、変更、追加、創作をしてはいけません。\n\nstoryboardDialogue:\n0～2s：【トラ＆マル】:"逃げろォォォ！"\nstoryboardAudioDescription:\n0～6s: 叫び声に重なる爆走의 足音、床を擦る紙と綿の連続音。\n6～10s: 壁に押し付けられるドン、直後にポンと戻る小さな弾み音。\n10～24s: ベッド下へ滑り込む擦過音、マルの体が詰まる布と床のギュッ、引っ張る摩擦。\n24～30s: 動きが止まっていく無音寄り、最後に小さな息だけ