Video Prompts for Anime

A woman with long black hair dancing in a cozy room. When she makes a frame, triangle, or square with her hands, an overlay appears inside. It starts as an anime-style avatar, transitions to a Spider-Verse-inspired comic book style halfway through, and finally returns smoothly to live-action.

Create a 30-second 16:9 chibi anime comedy short featuring two clearly adult male best friends, KAI and RYU, in a cozy apartment. No speech bubbles, no storyboard panels, no text overlays. Dialogue is spoken naturally with expressive voice acting, exaggerated facial expressions, physical comedy, and cinematic camera movement.\n\n0–5s — HOOK: Kai secretly sneaks behind Ryu while Ryu is confidently taking a selfie and whispers, “Bro… I just found your biggest weakness.” Ryu freezes dramatically: “My weakness?” Kai grins mischievously: “Your own confidence.” Quick comedic zoom.\n\n5–12s: Ryu challenges him, “Prove it.” Kai casually points toward Ryu’s phone. Ryu looks down and realizes the front camera has been recording him making ridiculous poses for the last five minutes. Kai bursts out laughing. Ryu tries to grab the phone: “DELETE THAT!” Kai runs away: “Never!”\n\n12–20s: Fast slapstick chase around the apartment. Ryu jumps over the sofa, Kai ducks under a table, cushions fly everywhere. Ryu finally corners Kai and mischievously says, “Got you.” Kai smirks: “Check your pocket.” Ryu looks down and discovers Kai secretly replaced his phone with a banana.\n\n20–26s: Ryu stares at the banana in complete disbelief. Kai is laughing uncontrollably. Ryu slowly raises the banana like a phone and says, completely serious, “Hello… Apple support?” Kai collapses laughing.\n\n26–30s — TWIST/PUNCHLINE: Ryu suddenly pulls out Kai’s REAL phone and reveals he secretly recorded the entire prank. Kai's smile disappears. Ryu grins: “Now we're even.” Both laugh and fist-bump as the camera rapidly zooms out.\n\nVisual style: polished colorful chibi anime, oversized expressive eyes, exaggerated reactions, playful body language, dynamic squash-and-stretch animation, detailed cozy apartment, cinematic depth of field, smooth character motion, energetic comedic timing.\n\nAudio: natural spoken dialogue, playful background music, footsteps, sofa impacts, phone notification sounds, exaggerated comedic SFX, short silence before punchlines, synchronized lip movement.\n\nPacing: extremely fast opening hook, escalating physical comedy, clear dialogue, three comedic reversals, strong final punchline. Keep everything wholesome, cheeky, mischievous, and genuinely funny.

Cinematic anime short film clip, 15 seconds. Tropical beach at midday, clear bright blue sky, white sand, ocean in background. ONE single thick tall bamboo tree standing in the sand — this is his only training target the entire video. All kicks and punches

Create a premium cinematic 15-second transformation story following the SAME young female protagonist continuously through three visual styles: realistic live-action → high-end anime/3D animation → Hollywood blockbuster. 0–5s — REALISTIC WORLD A stylish young woman walks alone through a modern city at night. Rain falls lightly. Neon signs reflect across the wet street. She wears a sophisticated black streetwear outfit with subtle red accents. She hears a powerful hip-hop beat. She stops. Close-up of her eyes. She looks toward camera and begins a confident dance-inspired movement. Typography appears: “WHO ARE YOU?” Camera rapidly pushes toward her hand as she raises it toward the lens. 5–10s — ANIMATED WORLD The exact hand movement triggers the transformation. The live-action city smoothly becomes a breathtaking high-end animated world. The SAME girl, SAME outfit, SAME pose and SAME movement continue without interruption. Stylized energy particles surround her. Buildings stretch upward. Neon colors pulse with the beat. She performs a powerful choreographed movement through the animated city. Typography dynamically forms around her: “WHO YOU CHOOSE TO BE.” Use fluid premium animation, detailed facial expressions, realistic cloth movement and cinematic camera choreography. 10–15s — HOLLYWOOD BLOCKBUSTER During her final animated movement, the environment suddenly transforms into an epic Hollywood-style cinematic cityscape. The SAME girl lands on a rooftop overlooking a massive glowing metropolis. Realistic wind moves her hair and clothes. Dramatic clouds, atmospheric haze, anamorphic lens flare and powerful backlighting create blockbuster scale. Camera circles her as she turns toward the lens. Final beat. She confidently smiles. Huge cinematic typography appears: “BECOME UNSTOPPABLE.” Final hero frame with dramatic music impact and premium movie-trailer finish. VISUAL RULE: The transformation should feel like ONE continuous shot/story—not three separate clips. Preserve the girl's identity, clothing, movement, camera direction and emotional expression through every style transition.

High-quality anime imagery. High-density 3D toon/cel-look cinematic high-speed swordplay set in a cyber-shrine. Theatrical-grade key animation, high-quality composite, transparent lighting, dense background art. Evasion, destruction, deflection, counter-attack, and final impact are depicted as a single continuous motion without resetting stances or idling. [References and Character Locking] image1 is the face reference and identity reference. image2 is the outfit reference and full-body reference, including accessories and weapon. If references conflict, face and identity have priority from image1; outfit, full-body design, accessories and weapon have priority from image2. image1 is the reference for the swordsman's face and character identity. Prioritize her soft and delicate small features, 17-year-old age feel, large purple-pink eyes, jewel-like irises, multi-layered white catchlights, thick upper and soft lower lashes, eye size and spacing, golden eye accents, light cheek flush, small mouth, bangs, long pink twin-tails, and purple flower and crystal hair ornaments. Do not use the background of image1. image2 is the reference for the swordsman's outfit, decorations, full body, physique, and weapon. Prioritize a slender silhouette about 160cm tall, dark blue and purple uniform-style layered outfit, gold trim, purple star-shaped brooch, white pleated blouse, purple ribbon tie, belt and waist decorations, purple layered pleated skirt and sheer layers, back ribbon, black tights, purple-black heeled boots with flower decorations, purple katana, diamond-wrapped hilt, star-shaped guard, purple tassel, and the scabbard on the left hip. Do not include image2's character sheet layout, borders, text, white background, or multiple poses in the video. If references conflict, prioritize face, eyes, irises, jaw, bangs, and hair color from image1; prioritize outfit, decorations, physique, silhouette, and katana from image2. Maintain the same person/character identity throughout the video. [Character A / Character 1: Swordswoman Fixed] Only one swordswoman. Do not change her facial features, eyes and irises, hairstyle and hair color, outfit and equipment, physique and silhouette, or the specific color scheme of pink, purple, and gold. Do not allow her to look like an adult, a generic face, or an expressionless Noh mask even during high-speed action or wide shots; maintain both eyes, irises, eyelashes, cheeks, and golden eye accents. The only things that may change are expressions, gaze directed at the enemy, mouth, posture, breathing, and natural hair/clothing delays. Do not hide the face with hair or effects. [Character B / Character 2: Cyber-Oni Sentinel Fixed] Only one Cyber-Oni Sentinel. Do not change its heavy black metal frame, horned mask, two metal claws on both forearms, cyan plasma exhausts, mass (approx. twice the swordswoman), and forward-leaning silhouette. It moves by shifting its center of gravity without slipping, performing horizontal slashes, deflections, pushes, and claw thrusts. Do not change the enemy into a crystal statue, crystal doll, mist, or abstract object. Do not mix traits between Character A and Character B. No face-averaging, outfit swaps, cloning, or extra characters. [Style Fixed] Rendering is 3D toon/cel-look only. Thin and delicate blue-violet colored outlines. Clear 2-3 step cel shading and transparent mid-shadows on face, hair, outfit, and weapons. Key lights and glows use the character's pink/lavender/violet palette; cyan neon is restricted to background/enemy accents. Background is one step darker than the character. High-density highlights, hair, lashes, embroidery, gold trim, ribbons, and jewels. Distinctly depict textures for skin, cloth, leather, metal, wet tiles, and glass. Backgrounds also maintain movie-level density. Avoid thick outlines, flat shadows, generic faces, or mixed styles. 3D toon/cel-look only: fine blue-violet colored linework; two-to-three-step cel shading; a pink, lavender, violet and gold character color palette; layered eye and hair highlights; high-density fine detail and high-density background detail; distinct material roughness for skin, cloth, leather, polished metal, wet roof tiles and glass. No thick outlines, flat single-layer cel shading, generic 3D anime faces, smooth plastic CG, low-detail backgrounds, photorealism, semi-realism, dull colors or style mixing. [Weapon Dimensions, Structure, and Operation Fixed] The swordswoman has five katanas total: one handheld (right hand) and four floating katanas operated by telekinesis. Only one sheath on the left hip. All five must be Japanese katanas with the same dimensions and material as image2. Each katana is approx. 95-105cm total length (70-75cm blade, 23-28cm hilt). Polished dark purple metal blade, straight edge, hamon, tip, star-shaped guard, diamond-wrapped hilt, and pommel. Show the full blade clearly as the primary subject. Four floating katanas start at fixed asymmetrical points behind her (high left shoulder, higher right shoulder, left hip, lower right knee). Keep clear space between the character and swords. She controls them via gaze and finger movements. Glowing effects are limited to fine purple light along the edge/hamon. Linear purple/magenta afterglows appear only during movement. No daggers, light-sabers, or glowing toys. In combat, launch one by one while others remain stationary at their waiting points. Each launch follows a straight path to a single contact point, then a new straight path to a stop. Swords move as rigid bodies. No orbiting, rotating shields, or propellers. During the final maximum collision, all four are used once, arriving from different depths/times. They do not merge or form a cross. Katanas do not multiply, disappear, or penetrate the character. Right hand grips only the hilt behind the guard. Sentinel's weapons are two metal claws integrated into forearms. No katanas for the Sentinel. [Stage] Rooftop temple district at night. Connected wet tile roofs, scaffolding, neon torii, cables, and steel pillars. Thin steam and reflections. Tiles and pillars serve as depth markers. Use close-ups and medium shots to keep character and sword dimensions readable. [Combat Motion Contract] Body mechanics remain readable: supporting leg, center of gravity, hips, shoulder and footwork physically drive every move. Every parry or impact shows a visible pre-contact gap, one named contact point, then reaction force, recoil, hit-stop and sparks only after contact. Continuation uses the momentum and reaction force into the next attack or counter without a neutral reset. [Continuous Action] Start: Swordswoman on mid-ground right facing enemy on left, supported by right leg. Handheld katana in right hand. Four floating katanas stationary behind. Enemy steps in from foreground left. When the enemy slashes, the swordswoman ducks and moves forward. Expression changes from tracking the claw to focused gaze. Roof remains intact until contact; then sparks/tiles/steam burst diagonally. Swordswoman pushes forward, twin-tails and ribbons flowing with a delay. One floating katana launches from lower right knee toward the enemy's claw. Clear gap before contact; single contact point with sparks and hit-stop. The deflected sword hits a steel pillar, which only breaks after contact. The second floating katana from high right shoulder strikes the other claw, opening the defense, then stops at a point upper right. Camera follows low from right to left. Background axis is diagonal. [Clear Mid-Combat Exchange] Enemy turns back and pressures with right claw. Swordswoman parries once with handheld katana while sliding her left foot back. One floating katana from high left shoulder launches straight at the enemy's forearm/claw. It hits, causing sparks and opening a path. Swordswoman passes through and slashes the armor joint once. Causality: pressure -> parry -> floating sword interception -> opening -> character slash -> enemy knee sink. Camera stays mid-shot at hip height. [Final Counter and Maximum Collision] Enemy counters with left claw; swordswoman evades. She slashes the armor joint with the handheld katana while the four floating katanas launch from their points, hitting adjacent spots with slight delays. They do not merge or spin. Shockwaves and pressure surfaces arise only after contact. The enemy is blown back, crashing through a torii into a steel core which buckles. Camera shows the result in the same frame as the swordswoman's face. [Audio] No BGM. Synced foley: boots on tiles, servo sounds, wind, metal clashing, tiles breaking, armor falling, and steel buckling. Exhalations only at key moments. No dialogue. [Important Fixed] Characters are one swordswoman and one Cyber-Oni Sentinel. Katanas consist of one handheld and actual-size flo

Create a short anime-style cooking video showing hands preparing spicy cheesy ramen: pick instant spicy ramen from a wooden pantry, place a small pot on a gas stove, pour in water and boil the noodles over vivid blue flames. Add dark red spicy sauce and a slice of yellow cheese, then stir with chopsticks until creamy and glossy. Pour the ramen into a white bowl, then serve it on a wooden table with chopsticks and a red can of cola. Cozy, cinematic lighting, detailed food textures, smooth close-up shots, and satisfying cooking sounds.

Low camera angle following a girl pedaling her bike down a gentle slope facing a scenic bay and green mountain. Gentle motion blur on wheels, realistic wind blowing her blonde hair, sparkling ocean water in the background, Studio Ghibli-inspired vibrant lighting. (Immersive

Mio uses the glass-attached version in the first half -> virtual form in the second half (hair tips glow cherry-colored, no glasses), Yoru uses the glowing form. The appearing characters are drawn in exactly the same appearance and color scheme as the attached reference materials. A 10-second scene from a Ghibli/Summer Wars-style hand-drawn 2D cell anime theatrical movie. Soft outlines, pale watercolor background art, cell-style 2-stage shading, and rich volumetric light. Starts from the continuation of the glass rim lighting up in the previous scene. [Most Important] Start with a faint blue light from the glasses at 0 seconds, and within 2 seconds, the blue-gray reality dissolves into shards of glowing glass, beginning the fall into a richly colored virtual world. This visual transition is the highlight of the work. 0.0–1.0s: Super close-up of the girl's eyes. Girl Mio, 14 years old. Short black bob, bangs cut straight above the eyebrows, star-shaped silver hairpin on the right temple. Large amber eyes, light freckles on cheeks. White short-sleeved sailor dress, navy blue lines on the collar, navy blue ribbon at the chest, knee-length skirt, red sneakers. A ring of light runs through the smart glass lens, and the rich colors are reflected in her eyes. 1.0–2.0s: The world ripples like the surface of water, and the blue-gray shopping district quietly dissolves into countless shards of glowing glass. Depict it as transformation rather than destruction. Mio's body is softly lifted into the air. 2.0–4.0s: Mio falls through a pure white infinite space. Colorful rivers of light spiral around her. A tail of light appears in front to lead the way. Virtual world Yoru's outline is bordered by amber light particles, the tail pulls a meteor-like band of light, and the bell lights up in amber. 4.0–6.0s: The fall turns into flight. A full view of the virtual world below: groups of white floating islands, colorful rivers of light, and countless colorful avatars. The camera follows Mio from behind at super high speed. The virtual world Mio has the same appearance, but the tips of her hair glow pale cherry color, and the smart glasses have disappeared. 6.0–8.0s: Piercing through the river of light alongside Yoru. Colorful bands of light pass by at high speed on both sides, and Mio's hair and skirt flutter in the wind. 8.0–10.0s: Rapidly rising toward the giant white tower towering in the center, and the moment she crosses the top, a super wide-angle shot opens up the entire virtual world below. Mio laughs aloud. Ends with a sense of liberation. Audio: Startup sound at 0s, clear sound of glass shards at 1s, and three-dimensional sounds of wind and light particles during the fall. The music features a rising harp glissando, strings rapidly rising at 2s, full orchestra at 6s, and open chords at 9s. Do not use destruction/collapse expressions for the real world (depict it as a transformation into light). No text or subtitles. Live-action, photorealistic, and 3DCG are prohibited.

Use @dargo as the main character, keeping facial features, hairstyle, and body proportions consistent throughout. He is a 25-year-old man. Cinematic time-freeze anime short film, 18 seconds, ultra-realistic 2D anime character composited into a realistic 3D setting, 35mm lens, moody sports bar interior lighting with neon accents, volumetric haze, dynamic hard shadows, shallow depth of field. [0:00–0:04] A packed, dimly lit sports bar, time flowing normally. Steadicam frontal medium shot tracking him as he walks confidently through total chaos with a smug smirk. He is clearly a 2D anime-styled man inside a realistic 3D bar environment, creating a striking mixed-media contrast. On the blurred TVs in the background, the Yankees have just clinched the World Series. The crowd erupts in euphoric celebration. As he walks, he slowly raises his right hand, snaps, and shouts, “ZA WARUDO.” [0:04–0:07] At the snap, a subtle spherical shockwave bursts from his fingertips with air distortion and light refraction, expanding outward through the bar. Everything freezes mid-eruption. Thick golden arcs of beer suspend in the air with sharply visible droplets. Popcorn kernels float motionless. Fans are frozen mid-cheer, mid-jump, and mid-embrace. Neon highlights catch the suspended liquid, dust, and haze. The entire room falls into absolute silence. [0:07–0:10] Only he moves. His footsteps echo faintly in the frozen bar. The camera glides backward as he walks through the suspended celebration, calm and amused. He casually ducks under a frozen splash of beer, then notices a cluster of floating popcorn. He reaches out and plucks a single kernel from the air with total ease. [0:10–0:13] He stops in front of a frozen woman in a Yankees jersey, her face locked in an ecstatic, tearful scream, arms raised for a high-five. He tilts his head with quiet amusement, gently adjusts the brim of her frozen Yankees cap, then gives a small approving nod. [0:13–0:18] He flicks the popcorn kernel into her open mouth, turns toward camera, and smirks. He snaps again. A stronger reverse shockwave bursts outward and restores motion instantly. The beer completes its splash, the crowd’s deafening roar slams back in, and fans land from their jumps. He continues walking away through the wildly celebrating crowd as the camera pushes forward through the bar. Fade to black. Sound: deafening sports bar celebration and TV crowd noise → finger snap → deep shockwave bass drop → absolute silence → faint footsteps and room tone → small popcorn flick and crunch → second snap → reverse shockwave → full crowd roar returns instantly.

Task: Dreamina Seedance 2.5 — Omni ReferenceDuration: 30 secondsAspect ratio: 16:9Editorial target: 24fps, F000-F719 [Generation Goal] A high-speed parkour tag game featuring only three characters from @Image1, @Image2, and @Image3. Blue Character C is the 'it' and the target fleeing. Character A and Character B chase Character C. Never use a composition where C chases A and B. Throughout the entire video, C must lead while A and B chase from behind. Use sharp movements like martial arts positioning, feints, and evasion, but no physical attacks like punching, kicking, grabbing, or body slamming. Characters never make physical contact. No slow motion, freeze frames, long pauses, or static poses. Maintain normal or high speed throughout. Place rapid accelerations and direction changes (~0.2s / 5 frames) at key moments. Connect these directly to running, landing, and the next step without stopping. When C stomps, the compression to extension of legs, hips, and shoulders links up, and C immediately dissolves into a group of liquid brushstrokes to move instantly. Generate a short camera shake of 6-8 frames or less only at the stomp point. Do not make muscles, joints, feet, ground, or points of force glow. Do not generate internal light, glowing muscles, auras, electricity, magic circles, or glowing footprints. Do not destroy the ground or equipment. [Material Roles] @Image1 = Character A (Chaser). Maintains yellow bob hair, cat ears, green eyes, light blue/white clothes with yellow trim, triangle tie ornament, pink backpack, pink boots, light blue tail with yellow triangle tip. Agile and bold. Chases while watching C's back and landing points. Expressions: smug smile, surprise, focus, and excitement when catching up. @Image2 = Character B (Chaser). Maintains lime green round bob, two leaf-like ears, golden eyes, white puff sleeves, pink bowtie, dark green shorts with suspenders, green boots, and a long lime green tuft from the back of the head with a round tip. Predicts C's escape route and flanks from the outside. Expressions: observation, discovery, focus, surprise, and competitive fun. @Image3 = Character C (Blue 'It' / Escaper). Maintains vivid blue long hair in a ponytail, two white horns, pointed ears, golden eyes, star earrings, dark red Japanese attire, black skirt, black belt/back ribbon, black knee-highs, and red geta. Always runs ahead of A and B. Checks chasers over the shoulder and uses obstacles to escape. Expressions: relaxed smile, surprise when almost caught, and smug smile when escaping. Do not treat three-view diagrams as clones. Only three people total. Do not swap faces, hair, costumes, ears, horns, tails, backpacks, shoes, or color schemes. Do not generate weapons. @Image4 = Park background. Use only the structure, equipment, color tone, depth, and spatial relationships. Do not reference people. [Environment Layout] Maintain the @Image4 park as a single continuous space. Features: wooden bench at front-left; large tree, sign, and streetlight on left; gazebo with green roof and branching path in center; wooden arched bridge, pond, rocks, lilies, and rope fence at center-right; swings, slide playground set, sign, and weeping willow on right. Path order: Bench -> Tree/Sign -> Streetlight -> Gazebo -> Arched Bridge -> Pond/Stones -> Swings -> Slide set -> Weeping Willow -> U-turn -> Far side of pond -> Front of gazebo -> Front-left walkway. Do not change this order. Do not move equipment between cuts. Do not teleport characters. Carry over the end position of each cut as the start position for the next. [Chase Direction Lock] F000-F440: All three move from screen left to right. Character C leads at the far right. F441-F460: C performs a hairpin turn using the weeping willow as a pivot. Reverse direction only within this cut. F461-F719: All three move from screen right to left. Character C leads at the far left. Do not reverse left/right without reason. Do not cross the imaginary line. [Continuous Action Rule] Do not have characters waiting at the start of a cut. Start from the middle of the action from the previous cut. A landing foot connects to the next step. A hand placed on the environment connects to pushing the body off. Direction changes happen immediately after eye movement. Maintain these movements even in close-ups. Do not switch obstacles mid-action. Inherit running speed, horizontal/vertical positions, and body orientation between cuts. [Unstable Line Rule] Render the entire video as 'continuously generated lines and watercolor'. Redraw each frame as a new rough sketch. Do not create a stable outline. Leave 30-60% of the body perimeter as white paper gaps. Lines should be thin/broken, suddenly thicken, shake, branch out, or vanish. Watercolors should bleed outside outlines, leave white spots, show dry brush marks, or be misaligned. Body movement leads, pencil lines lag by ~1-2 frames, and watercolor surfaces lag further behind. In slow movement, draw ~60% of the body. At medium speed, draw ~30-40%. At high speed, draw ~10-20%. At max speed (~0.2s), represent characters with only 3-5 brushstrokes and 2-3 rough watercolor planes. Speed forms: A = Yellow/Light Blue/White/Pink; B = Lime/Dark Green/White/Pink; C = Blue/Dark Red/Black/Gold. Do not just add speed lines to a finished body; treat the liquid brushstroke groups themselves as the moving character. [Cuts 1-20] (Details the specific choreography through the park from Cut 1 starting at the bench through Cut 20 ending in a dynamic, unresolved high-speed chase in the front-left walkway, following the established environment sequence and fluid movement rules). [Camera Treatment] Create a sense of speed with low horizontal tracking and foreground objects (pillars, fences, branches, swing chains, bridge rails) passing the camera. Do not follow character fronts for long. Focus on side views, diagonal rears, and low tracking. Use wide or full shots to clearly show the C-in-lead / A/B-behind composition. Limit close-ups of eyes/feet/hands to 0.2-0.5s cutaways that connect directly back to the same action in wide shots. Camera shake only on C's acceleration stomps. [Expression Direction] Character A: Focused eyes targeting C's back, smug smile when close, surprise when C pulls away, high excitement. Chara

High-quality anime footage. Theatrical-class high-density 2D hand-drawn action animation. High-speed melee combat. Set in a dark near-future giant corridor, the protagonist breaks through multiple enemies continuously without stopping. The most important elements are overwhelming speed, grounding, reaction force chains, clear contact, extreme camera distance changes, brief 'face rewards' (close-ups), and momentary graphic impacts in red, black, and white in the second half. No static repositioning phases. Each attack's reaction force must be used as the propulsion for the next movement or attack. [Character Consistency] Maintain the protagonist from the source reference image as the same person across all cuts. Do not change face shape, age, eye size/spacing, iris color, jaw, bangs, hairstyle, hair color, outfit, body type, or specific color scheme. Do not allow the character to look like someone else, appear more adult, or be simplified even during high-speed movement, distant shots, or backlit explosions. Accurately restore the face, eyes, and hair to match the source image the moment the camera returns to a close-up. The protagonist's weapon is a single slender single-edged sword. Maintain the blade, guard, and hilt as a single continuous structure. Do not multiply, split, or exchange weapons. One or both hands must always grip only the hilt; do not overlap the blade and hands in front of the face. [Enemy Consistency] 4 humanoid enemy soldiers wearing black near-future armor. Maintain the same armor design for all 4; do not mix face, hair, outfit, or weapons with the protagonist. Enemies maintain humanoid bodies, black armor, and limbs during combat. Do not transform into crystals, abstract objects, or light masses after being hit. [Art Style Consistency] 2D hand-drawn animation throughout. Thin delicate line art, dark but not overly black outlines, 2-3 layers of anime shading, transparent mid-tones, multi-layered highlights on hair and eyes, high-density background art, exaggerated key animation for theatrical action. Use short smear frames with boldly stretched shapes for high-speed motion where necessary, but return the face to clarity during close-up frames. No thick black outlines, simplified TV animation, low-density backgrounds, flat single-layer cel shading, smooth 3DCG, live-action, semi-realistic styles, dull colors, or mixed art styles. [Setting/Color/Light] The setting is a dark, massive near-future corridor. Maintain the same location throughout. Background colors: black, dark navy, dark teal. Thin red linear lights on walls, floors, and ceilings pointing toward the vanishing point to emphasize speed and depth. Fixed color roles: Cyan-Blue = protagonist's acceleration and electrical afterglow; Lime-Yellow-Green = sword attack trails; Orange-Yellow = sparks and explosions after contact; Red-Black-White = graphic impacts for maximum shock only in the second half. Do not swap color roles between cuts. Stabilize skin tones and protagonist colors; no random white balance changes. [Common Movement Rules] Do not process running as a parallel slide. Sequence: grounded foot fixed to floor -> pelvis moves past grounded foot -> opposite knee moves forward -> toe of grounded foot kicks off -> short aerial phase -> opposite foot lands forward -> knee compression -> next kick. Pelvis and thorax must move forward past floor joints. Movement in world coordinates must not be erased even if the camera follows. Landings must pass through: foot contact -> knee compression -> hip sink -> upper body follow-through. Sword reaction force flows from sword -> wrist -> elbow -> shoulder -> thorax -> hips -> grounded leg to propel the next move. Hair and clothing follow the body with a slight delay. [Contact Rules] All key attacks follow: clear gap before contact -> contact at specified position -> very short hit-stop at contact -> sparks, armor displacement, and explosion only after contact. Do not blow back enemies or generate explosions before contact. No static screens showing only effects; continue character movement during sparks/explosions. [10 Cuts Sequence] Cut 1: Floor Ignition. Low-angle close-up at feet. Lead foot grounds strongly, toe-kick generates lime-green sparks. Cut 2: Corridor Acceleration. High-speed strides toward vanishing point. Wide-angle perspective. Cut 3: Face Reward to Strike. Quick camera approach to face. Diagonal upward slash with hit-stop and lime trail. Cut 4: Explosion & Axis Change. Enemy hits equipment causing explosion. Protagonist steps around smoke to find next enemy. Cut 5: Continuous Melee. Parry then horizontal slash on second enemy. Move past to third. Cut 6: Spatial Opening. Jump over third enemy's attack. Wide long shot to show the massive corridor scale. Cut 7: Landing to Re-acceleration. Grounding sequence, cyan afterglow, and quick face reward during next dash. Cut 8: Third Contact & Max Pressure. Torso slash on third enemy. Extreme close-up on eye leading to graphic insert. Cut 9: Graphic Impact Intercut. Flash of red/black/white 2D graphic frame to emphasize impact. Immediate return to reality. Cut 10: Final Impact. Approach fourth enemy without stopping. Diagonal slash with strongest hit-stop and red/black/white graphic frames. End on the motion of the next step; no victory pose or static shots. [Sound] High speed but varying volume. Sync breaths and metallic sounds with action. No long screams. [Important Guards] No neutral resets; every force becomes propulsion. Move character relative to floor, not just the camera. Vary camera distance/angle/size between cuts while maintaining movement vector. No text, logos, clones, or sudden enemy additions. No long slow-motion, stalemates, or final victory pull-away shots.

A Japanese anime-style 2D animated close-up video. Hand-drawn anime hands press vibrant green pandan rice flour dough through a traditional press into a bowl of liquid nitrogen, creating a whimsical cloud of cold steam and sparkling vapor

Setting: A cozy indoor room with warm, ambient lighting, featuring posters on the wall, a hanging lamp, and modern furniture. Subject: A young woman with long dark hair, wearing a gray hooded jacket, performing a rhythmic, choreographed hand and body motion dance in front of the camera. Key Visual Effects: Animated Framing & Transitions: As she moves her hands to form frames and geometric shapes (triangles, rectangles) in front of her face and upper body, colorful animated overlays appear inside the framed spaces. Anime & Superhero Styles: The initial overlays display an anime-style digital avatar matching her expressions. Midway through, the overlay switches to a Spider-Gwen / Spider-Verse comic style, complete with a hooded mask and neon cityscape backgrounds. The animation dynamically shifts and transforms along with her hand movements, smoothly blending back into real life by the end of the clip.

recreate it with an X-Man version in picture form, make the character say: “Basically, the goal is to give people hope, realistic hope. Maybe hope is even the wrong word. Simply to say that there is a path to a fully sustainable, global economy. We are on that path, and we are accelerating that path. So long as we don't get complacent about it, it will happen.”

Create a 30-second vertical 9:16 cinematic cooking short using all uploaded reference images. REFERENCE PRIORITY: Character sheets = exact character appearance. Kitchen reference = exact environment. Pancake recipe image = food preparation and final pancake appearance. VISUAL STYLE — CRITICAL: Create a deliberate mixed-media scene. The YOUNG WOMAN must be ULTRA-REALISTIC LIVE ACTION: photorealistic face, skin, eyes, hair, clothing, natural body proportions, realistic expressions and physically accurate lighting. She must look like a real human filmed with a cinematic camera. NOT anime, NOT cartoon, NOT illustration. The FOX GIRL and FOREST GUARDIAN must remain 2D FLAT ANIME-STYLE STICKER CHARACTERS based exactly on their uploaded character sheets. Preserve their clean illustrated outlines, flat colors, cel-shading, stylized proportions and original designs. They must NOT become realistic, 3D CGI or live-action characters. The kitchen is photorealistic. The realistic woman and realistic kitchen contrast intentionally with the colorful 2D anime sticker characters. Make the 2D characters feel naturally integrated into the real kitchen through correct scale, perspective, contact shadows and lighting, while clearly remaining flat illustrated characters. CHARACTERS: Young woman: use uploaded reference exactly; long black hair, white athletic jersey, black shorts, white socks and sneakers. Fox girl: use uploaded reference exactly; pink hair, fox ears, blue eyes, fluffy cream/orange tail, purple/magenta outfit. Energetic, playful and slightly clumsy. Ears and tail react naturally. Forest guardian: use uploaded reference exactly; huge antlers, wooden body, green leafy beard/mane, wooden armor, blue glowing eyes/rune and living vines. Calm, serious and gentle. His vines act as extra hands and use harmless nature magic. LOCATION: Photorealistic warm modern kitchen, wooden counters, cream cabinets, stove, island and soft afternoon sunlight. Keep the same kitchen, lighting and layout throughout. STORY + TIMING: 0–5s: Young woman walks naturally into the kitchen. Smooth cinematic tracking shot from behind, then over-the-shoulder reveal. She sees the 2D fox girl and giant guardian preparing pancakes at the counter. GIRL: "What are you two doing in my kitchen?" FOX GIRL: "Pancakes!" GUARDIAN: "Pancakes." Girl blinks: "...Okay." 5–10s: Girl joins them. GIRL: "Fine. You're helping." They begin making pancake batter following the uploaded recipe: flour, sugar, baking powder, baking soda, salt, milk, egg, melted butter and vanilla. Guardian's vines naturally hold ingredients and utensils. 10–15s: Fox gi

Create a cinematic 16:9 anime fashion short using the uploaded character sheet as the exact character reference. Keep the same blonde female cowgirl, black western hat, sunglasses, cropped black top, denim shorts, boots, accessories, facial features, hairstyle, and proportions consistent throughout. Set the scene on Route 66 at golden-hour sunset. A black vintage muscle car is parked beside her on an empty desert highway. 0–4s — She leans casually against the car, adjusts her hat, then looks toward the setting sun. Warm wind moves her long hair naturally. 4–8s — She turns and slowly walks down the highway as the car drives away into the distance. Wide cinematic composition, orange sunset, desert landscape, strong backlight. 8–12s — Low-angle tracking shot of her boots walking along the road, followed by a close-up as she adjusts her sunglasses. Preserve outfit and character details exactly. 12–15s — Tight cinematic portrait. Sunset reflections appear in her sunglasses, hair moving softly in the wind. She looks toward the horizon with a calm, confident expression. End on a polished anime film hero frame. Style: premium anime PV, western fashion editorial, Route 66 road movie, cinematic sunset lighting, warm orange and black palette, detailed character consistency, natural hair and fabric motion, shallow depth of field, smooth camera movement, no character changes, no distorted hands, no extra accessories, no subtitles or watermark.

High-quality anime footage. High-density 3D toon/cel-look. A cinematic high-fantasy action set in a magical city after the rain. A pink-haired girl mage intercepts the fiery breath of a giant black dragon with a defensive array from her grimoire, uses the reaction force to teleport to the side, fires a high-output beam from a multi-layer convergence gate in her right palm, and causes it to collapse into black ash and magical particles. [Character Identity Fixation and Role of Reference Images] image1 is the primary reference for the protagonist's face and identity. Maintain a soft oval young face, small chin, large blue-violet to purple-pink eyes, multi-layered irises with white catchlights, delicate eye area, long coral pink twin tails, light bangs, lavender flowers on both sides, black-purple ribbons, and gold decorations throughout. image2 is the reference for the protagonist's costume, decorations, full-body design, and equipment. Maintain a slender and petite build, dark blue to purple gold-rimmed cloak, white frilled blouse, black-purple corset, purple multi-layered pleated skirt, white knee-highs, dark purple boots with metal fittings, waist belt, amethyst crystals, magic potions, and a grimoire with gold decorations. If references conflict, prioritize image1 for the face, eyes, irises, bangs, hair color, and hair ornaments, and image2 for the costume, decorations, build, equipment, and full-body silhouette. The two images are for appearance reference only; do not bring backgrounds, compositions, poses, white backgrounds, setting sheet borders, text, or split layouts into the video. Only concentrated expressions, eye direction, breathing, combat stances, and natural movements of hair, cloak, and skirt may be changed. Do not generate facial averaging, feature mixing, costume swapping, equipment multiplication, alter egos, duplication, or additional characters. [Art Style Fixation: High-Density 3D Toon/Cel-Look] This style fixation does not override character identity fixation. Thin, delicate colored outlines. Clear two- to three-level cel-shading and transparent intermediate shadows on the face, hair, costume, and equipment. Multi-layered highlights on eyes and hair. Use pink, lavender, dark blue, purple, and gold as the protagonist's color axis, and do not muddy skin tones. Differentiate fabrics, frills, leather, metal, amethyst, wet stone, and glass with different reflections and roughness. Movie-quality key animation, high-quality compositing, transparent lighting, and high-density background art. Only for magic, use high-brightness graphic animation VFX, volumetric light, additive emission, refraction waves, and directional motion blur. Avoid thick black outlines, flat single-layer cel shadows, generic 3D bishojo faces, smooth plastic CG, low-density backgrounds, semi-realism, realism, dull colors, or mixed art styles. [Protagonist's Equipment and State Transitions] Only one grimoire. Maintain the same binding and thickness with gold decorations throughout. Open it with the left hand only in Cut 1, close it and return it to the left hip holder in Cut 2, and do not open it thereafter. Keep amethyst crystals and magic potions attached to the belt. All attacks must come from the right palm, not from the grimoire or magic potions. Do not introduce items not in image2 or a second book. [Black Dragon Fixation] The enemy is a single giant black four-legged wyvern. About four times the height of the protagonist. One head, two horns, two glowing bluish-white eyes, a connected neck and torso, four legs, and three claws on each foot. One pair of wings, two in total. Each wing consists of wing bones and membranes continuous from the base of the back. One tail, continuous from the back of the waist to the base, scales, and tip without branching. The impact point is a single spot on the center of the chest scales. Do not create extra heads, horns, legs, wings, or tails. [Setting] A huge magical city after rain. Grand avenues of wet obsidian and blue-gray stone slabs, shallow puddles, old cracks, magical metal railings, and crumbling arch bridges. Surrounded by magic towers stretching to the heavens, spires, sky corridors, floating turrets, giant crystal pillars, magic lamps, glowing magic waterways, and layers of low mist. City lights are cyan, royal blue, violet, purple, and a small amount of magenta. Background light is one step darker than the protagonist, making her inherent colors of pink, lavender, purple, and gold the focus. No readable magic script; only abstract geometric patterns on the stone floor, crystals, and architectural surfaces should glow faintly. [Space and Continuity] Cuts 1 to 2: protagonist in the left foreground, black dragon in the right background. The attack axis is a diagonal line from right-back to left-front. In Cut 2, swap the left-right relationship only once intentionally during the teleportation through the magic circle passage. Cuts 3 to 6: protagonist in the right foreground, black dragon in the left background. Connect the right palm, the center of the convergence gate, the beam, and the center of the dragon's chest in a straight line. Do not swap left-right relationships from this point on. In adjacent cuts, at least two of the following must change: camera height, distance from the protagonist, character size in the frame, or background vanishing axis. Move the camera to show the point of contact, direction of force, face, or spatial relationship; do not substitute camera movement for the character's own movement. All movements should have natural acceleration, deceleration, inertia, and parallax. Cut 1: Fire Breath and Defense Array. The black dragon is in the right background. It steps its front legs onto the stone floor, pulls its neck back, its chest swells, and the inside of its throat glows orange-purple. The protagonist is in the left foreground. She opens the grimoire with her left hand and thrusts her right palm forward. Over the open page, a three-layer defense magic circle with a white core, cyan outer light, and purple-pink edges deploys front to back, rotating in opposite directions at different speeds. At the same moment the dragon swings its neck down, the orange-purple fire breath travels straight from the right background to the left foreground, evaporating puddles on the floor. There is still a clear gap between the tip of the fire breath and the defense array, keeping both the array and the protagonist unharmed. A short hit-stop occurs only at the moment the tip of the fire breath contacts the foremost array. Only after contact do the three layers of the array deflect in order from the back, patterns glow in a chain from the contact point outward, and the fire stream splits left and right to flow backwards. Reaction force returns in order to the palm, wrist, elbow, shoulder, waist, and back leg, with the back leg's sole stopping after scraping the wet stone floor. The twin tails, flower ornaments, cloak, and skirt flutter backwards with a time lag. The camera moves close to the protagonist diagonally from behind at a low angle along the axis of the fire breath, keeping the defense array, both hands, profile, and dragon's head in the same frame. Cut 2: Reaction Force to Teleportation. The protagonist pushes back the deflected foremost array with her right palm, channeling the reaction force into a rightward rotation of the waist to connect to the next action. The left hand closes the grimoire and returns it to the holder on the left hip. At the same moment it closes, the pushed-back array stretches into a large circular arc, crossing in front of the lens to become a complete screen-covering wipe. As the arc clears, the protagonist appears on a crumbling arch bridge to the left side of the black dragon. Rather than transparency, signify landing through ground contact of the soles, compression of the knees, sinking of the waist, and the cloak and twin tails catching up late. The black dragon, still in the posture of having finished breathing fire, begins to turn its neck and eyes toward the protagonist's new position. The camera changes axis during the full wipe of the arc, descending from a high diagonal rear to a low three-quarter view. Cut 3: Convergence Gate. Close-up of the protagonist's upper chest. Move in a gentle arc about fifteen degrees around the face from diagonally in front. Spread both legs, bend knees, and leave weight on the back leg. Point the right palm toward the black dragon's chest, fingers naturally spread. The left hand supports the outside of the right wrist. Show the connection of the palm, wrist, elbow, and shoulder; do not hide the entire arm with rings. In front of the right palm, a two- to four-layer incomplete concentric gate forms with a white luminous core, cyan outer light, and thin purple-pink edges. Each ring is separated front-to-back and rotates in opposite directions at different speeds. Short scales, radial lines, and blue-violet lightning gather at the firing axis; make the circles large enough to be cut off by the frame. The amethyst light of the palm pulses strongly twice, and the rings compress on the second pulse. Leave the face in the dark gap, not hidden by arcs. Expression changes in this order: first, the irises of both eyes brighten and narrow; next, the eyebrows move to the center; then, the eyelids narrow slightly; after that, the mouth is tightly closed; finally, the gaze is fixed straight on the center of the dragon's chest. Aim the tip of the nose, chin, chest, and shoulders at the same chest center. Cut 4: Long-lasting Maximum Output Beam. Low rear three-quarter wide. The camera follows at a low level just above the floor, advancing parallel to the beam, with the foreground magic railings, raindrops, floor magic patterns, floor reflections in puddles, and floating lamps flowing at different speeds. Emit a thick beam from the compression gate, occupying 35 to 45 percent of the outer diameter. High-density white to pale peach-white core, thick hot pink and magenta outer shell, violet intermediate layer, thin cyan-white edges, and blue-violet electric arcs. Do not shrink to a thin rod. Even after reaching maximum output, maintain the thickness, brightness, and axis continuously. Three compression waves run through the beam interior from the muzzle to the chest center in sequence; as each wave passes, the outer shell expands, pushing aside surrounding rain, mist, and water droplets. Do not cut the beam until all three reach the chest scales. Keep the dragon's chest scales, wing membranes, and tail unharmed until just before the first wave arrives; do not show cracks, glowing, or collapsing prematurely. Only at the exact moment the first wave contacts the scales at the center of the chest does that point indent. Only after contact do local sparks and scale creaking occur, front claws scrape the wet stone floor, and the impact transmits with a delay to the wing bases, membranes, neck, and tail. The protagonist endures the recoil without slipping her back leg, and the puddles on the floor ripple continuously in the direction of fire. Leave thick reflection bands of white-peach, magenta, and cyan on the floor. Cut 5: Beam Cessation and Giant Flamelike Magical Impact. In the same frame that the third compression wave reaches the chest, completely erase the entire beam length from the muzzle to the impact point. Do not leave thin afterglow. In that same frame, an incandescent core appears at the chest center, and a local hemispherical impact dome expands once. The dome immediately ruptures, and a giant flamelike aura erupts three-dimensionally upward and to the sides from the incandescent core. The aura keeps hot pink and magenta on the inside, violet and purple in the middle, and royal blue and cyan at the outer edges separated. Deploy sharp flame tips, thick volumetric light bands, transparent refraction waves, thin lightning, and outward speed lines at different depths and speeds in the foreground, midground, and background. Vary the size, angle, depth, and rise of each flame tongue. Integrate the beam axis that was straight until just before, the radially extending flame tongues, the circular impact ring, and the upward-rising turbulence all into the same impact point at the chest center, perceived as a single force. Limit pure white to around the impact core, leaving color outside the incandescent core. Maintain the state where high-saturation color planes of hot pink, magenta, and cyan occupy most of the screen for multiple frames before decaying, rather than ending with a momentary flash. Do not use full-screen whiteout. Multiple objects react at the same moment. Floor cracks run outward from the impact point in purple-pink, water splashes and steam spray outward, black scale fragments and stone shards scatter, the protagonist's hair is blown back strongly, surrounding pillars flicker, and the camera is pushed back. The camera makes a short diagonal dash toward the collision point, retreats slightly at the moment it receives the shockwave, and recovers with decaying shaking. The protagonist does not stop moving even in the blast. Keeping her right palm at chest height, she channels the impact through her palm, wrist, elbow, shoulder, waist, and back leg in sequence, her back leg's sole briefly scraping the wet stone floor to hold her ground, then passes that reaction force directly into the next stance, returning her left hand to the front of her body. Immediately after, gray-purple smoke masses and black scale fragments with size differences, dark cavities, internal purple light, vortices, and rising speed differences overlap in the foreground, completely hiding the black dragon's head, torso, legs, wings, and tail. Only the protagonist remains as a silhouette in the right foreground. At the end of this Cut, the protagonist maintains a low stance in the right foreground, keeping her body oriented toward the black dragon in the smoke, handing over that posture and position to the next Cut. Cut 6: Kneeling to Black Ash Disintegration and Combat Result. The camera moves to a medium distance over the protagonist's right shoulder with smoke gaps in the foreground, keeping the protagonist in the right foreground and the disintegrating black dragon in the left background in the same frame. The protagonist inherits the low stance, right palm height, and body orientation from the previous Cut. As the smoke center is split by the wind, the black dragon is kneeling with the chest impact point facing us. Both front legs are folded, only one front claw is on the floor, back legs are tucked, wings are half-closed, tail is extended on the floor, and the head hangs low. Do not return to a standing or re-attack posture. Amethyst and cyan-white magic cracks run from the chest center along the scale joints throughout the body. The moment the cracks reach the shoulders, the kneeling upper body lurches forward once, and short sparks scatter from the edges of broken scales. The edges of the black scales touched by the cracks carbonize and unravel into thin black ash, turning into fine purple, magenta, and cyan magic particles. Disintegration starts from the wingtips and tail tip, continuing through the wing membranes, wing bones, tail, limbs, torso, neck, and head. Do not physically sever wings or tail; instead, they particle-ize as if holes were expanding from the outside to the inside of the connected outline. Black ash rises into an ascending vortex, and magic particles spiral briefly and lose light. Finally, the magic core in the chest, the two horns, and the bluish-white light of both eyes crumble into particles, and the flesh, shadow, and floor reflections disappear completely. No corpse remains. The protagonist does not transition to a follow-up attack, keeping her stance while channeling the reaction force from her right arm to her waist. The endpoint is a single frame of the combat result. The protagonist in the right foreground maintains the upper chest size from the midground, remaining at a distance where her face, eyes, hair ornaments, and cloak can be read. In the left background, only the wet grand avenue where the black dragon was remains, with disappearing black ash and purple-cyan particles. The protagonist lowers her right palm a few centimeters while remaining in a guard stance, her twin tails and cloak settling with her breath. The last vortex of black ash rises in front of the lens, and the Cut ends on the frame where it is clear that the vortex has passed the top edge of the screen and no shadow or floor reflection of the black dragon remains anywhere. Do not transition to a long victory pose or additional facial close-up. [Audio] No dialogue, singing, or BGM. Sounds must not occur before their corresponding contact or action. Cut 1 synchronizes in order the black dragon's low growl and deep inhalation, the thick burning wind sound of the fire breath, the hard magic surface resonance the moment the fire breath contacts the array, the wind sound of the split fire flowing left and right, the protagonist's short breath, and the friction sound of the sole scraping wet stone. Cut 2 features the sound of dry leather and paper from closing the grimoire, high magic resonance as the array stretches into an arc, a short spatial sound of teleportation, and the sound of the sole touching the ground and fabric rubbing upon landing. Cut 3 has the high magic resonance and low convergence sound of the rings, and the pulsing sound of the palm. Cut 4 sustains the thick continuous roar of compressed magic, synchronizing the scraping sound of hard scales and low-frequency magic impact sounds each time the three compression waves contact the chest scales. Overlay the friction sound of the front claws scraping the stone floor. Cut 5 synchronizes the burst resonance of the compressed magic field as the magic dome ruptures, the cracking sound of wet stone, outward wind pressure, water sounds, and the collision sounds of stone shards and black scale fragments with the moment of contact, placing the protagonist's short exhalation at the end of the fire maintenance. Cut 6 synchronizes the fine scraping sound of dry scales unraveling into ash, the wind rustle of black ash, and the high crystalline sound of magic particles with the disappearance of each part. Stop the low tones at the moment the final core disappears, leaving only the high-pitched sound of particles and the protagonist's quiet breath for a moment. [Major Error Prevention] No text, subtitles, logos, watermarks, or setting sheet UI. No additional characters, second magic users, or alter egos. No shortened, thinned, flickering, or mid-way decaying beams, no premature explosions, no long simultaneous display of beam and explosion, and no lingering lasers. No symmetrical, flat flower-shaped explosions, same-sized round smoke, or pastelization. No full-screen whiteout. Do not composite the black dragon in front of the smoke. Do not show the dragon before the smoke clears. No recovery, instant disappearance, transparency, remaining corpses, severed wings or tails, or multiple tails for the dragon. Do not make the protagonist's face a generic or expressionless surface. No halos above the head, light points sticking to the face surface, or thick frames crossing the face.

High-quality anime footage. Urban fantasy action RPG style, high-rarity 3D toon / cel-look anime fast-paced melee combat. Most important is high tempo, recoil transition from spear strike to tail strike, clear direct hit with the tail, and heavyweight knockback. No slow motion, waiting, staring, long stances, or long charging times. [Reference Usage] Use Image 1 as the sole reference for fixed character identity. Maintain the character in Image 1 as the same person, prioritizing face, eye shape, iris color, hairstyle, hair color, bangs, outfit, decorations, physique, full-body silhouette, atmosphere, and character colors. Do not reproduce the background, room, furniture, pose, camera angle, framing, text, setting sheets, or UI from the reference image itself. Only facial expressions, gaze, mouth, combat stance, breathing, and natural hair/clothing sway may change. Mixing features, averaging faces, swapping outfits, changing the person, cloning, or adding extra characters is prohibited. Depict the protagonist exactly as the person in Image 1. However, to establish the fixed scene, unify them as a melee combat character with one long spear and one tail. Naturally adjust the shape, surface texture, decorations, glow color, and small item designs of the spear and tail to match the character colors and atmosphere of Image 1. Metal parts of the outfit, accent colors, decorations, spear glow, tail scale color, and overall screen auxiliary colors may also be changed to match Image 1. Location and combat composition are fixed; the weapon is fixed as a long spear. [Fixed Protagonist] The protagonist is the same person in all cuts. Maintain the facial features, eyes, iris, hair, outfit, decorations, physique, silhouette, atmosphere, and characteristic colors from Image 1. Use only one long spear. Keep the tip, shaft, and butt as a single rigid body, gripping only the shaft with both hands. Do not rotate only the tip. Only one tail, connected to the center-back of the waist, thick at the base and tapering naturally toward the tip. Connect the rotation recoil of the whole body to the tail strike. [Fixed Enemy] The enemy is a single giant armored goblin. Maintain the grey-green face and skin, long pointed ears, low nose, two tusks from the lower jaw, small amber eyes, broad heavy physique, sooty black iron full-body armor, large pauldrons, leather belts, old bronze studs, short dark red loincloth, and one battle axe. Approximately 1.8x the height of the protagonist. On the right side of the black iron breastplate, there is a wide rectangular reinforcement plate with three large old bronze studs arranged vertically; fix this as the target for the tail strike. Maintain the physical reality of the flesh and armor until the end, with no crystallization, particleization, or disappearance. [Fixed Art Style] Use only high-density 3D toon/cel-look. Complete 3D model feel, thin clean anime outlines, clear 2-3 level cel-shading on face, hair, outfit, spear, tail, and armor. Eyes and hair have multi-layered highlights, skin is clean toon-matte, cloth has low reflection, metal has hard edge reflections, and the surface of the tail scales is readable. Theatrical-grade key animation, high-quality compositing, transparent lighting, and dense background art. Avoid thick black outlines, flat single-layer shadows, generic 3D anime girl faces, low-density backgrounds, realistic styles, or mixed art styles. [Stage and Lighting] The stage is fixed: the rain-slicked rooftop of a Neo-Gothic Great Library Tower overlooking a neon city. Wet black stone floor, low glass walls, spires, large metal ventilation equipment, distant skyscrapers, shallow neon bokeh. Do not change locations. Fine-tune light and color to match Image 1, making the protagonist's colors the star. Background should be low-saturation, prioritizing the visibility of contact points with the protagonist, spear, tail, and armor. [Action] Cut 1: Follow from low 3/4 rear of the protagonist, tracking close to the ground. As the goblin raises its axe, the protagonist takes one sharp step in and delivers a single large diagonal sweep with the entire spear. Short hit-stop at the moment the tip hits the center of the black iron breastplate. Metal sparks fly, and the goblin's upper body opens backward. The protagonist does not reset their stance, immediately transferring the spear strike's recoil into a full-body spin to link into the next tail strike. By the end of the cut, the wide scale surface of the tail's outer third is already positioned one fist away from the reinforcement plate on the goblin's right side. Cut 2: Fixed close-up dedicated to the tail contact. Place the goblin from chest to waist at about 70% of the screen. Do not return to the protagonist's face or spear. The stars are the goblin's torso, the right-side reinforcement plate, and the single tail entering from the left. Show the one-fist gap between the wide scale surface of the tail and the reinforcement plate at the start. The tail enters in a low horizontal arc, striking the wide scale surface (not the tip) strongly once against the plate. Show the space before impact, the approach of the scales, the single contact, and the armor's reaction in sequence within the same fixed close-up. 0.04-0.06s hit-stop at impact. Remove speed lines, freezing the contact boundaries of the scales, the three bronze studs, and the reinforcement plate. The plate dents deeply, the torso buckles sideways from the contact point, and is pushed out of frame by that force. Cut 3: High-speed whip pan in the direction the goblin is pushed, returning to a low-angle medium-distance view near the ground. The goblin is blown horizontally at high speed about 3 meters (at least one body length) across the wet floor due to the tail strike's impact. Both feet leave the floor only after impact. It does not jump and does not spin on its own in the air. The axe remains gripped in both hands. The back armor and pauldrons crash directly into the side panel of the large metal ventilation unit; the panel dents deeply, spraying a film of water and short sparks outward. All movement stops completely upon collision, leaving the goblin slumped in front of the unit as a physically incapacitated entity. The protagonist does not chase, absorbing the counter-force on the spot. Cut 4: After a water splash wipe, a 3/4 front diagonal medium-distance fixed shot of the protagonist. Head to feet and the entire spear are in frame. In the background, the dented ventilation unit and incapacitated goblin remain. The protagonist is completely stopped at the tail-strike point. Naturally shift weight to one leg, gazing toward the fallen goblin. The tail settles in a low natural arc behind the legs. The protagonist moves the spear to their side, gripping the shaft near the center with one hand, letting the other arm hang naturally. Align the tip straight up, shaft vertical, and butt straight down, briefly lowering the butt onto the wet black stone floor. Show the gap between the butt and the floor, the single contact point, and the small water ripples after contact in sequence. The final frame shows the protagonist completely still, spear standing vertically by their side, butt touching the floor, tail settled low behind them, and the goblin incapacitated in front of the dented equipment in the background. [Important Guards] Only one protagonist, one giant armored goblin, one long spear, one tail, and one battle axe. Cut 2 is an independent fixed close-up of the tail contact only. Do not blow the enemy away before contact. The goblin must not just flinch; it must be blown horizontally as a physical object and stop upon crashing into equipment. The protagonist must stop completely on the spot after the tail strike, with no forward sliding, side-slipping, knee-sliding, or approaching the camera. The final sequence must be a medium-distance fixed shot showing the head, both feet, the entire spear, and the butt's floor contact simultaneously. No clones, humanoid afterimages, spear duplication, axe duplication, or tail branching.

Goal: Generate a viral-potential short dance video featuring the protagonist from the reference video. The subject is a single girl. The main event is a 30-second cute dance performance switching between 5 camera angles (front, top, left, right, rear) every 2 seconds, maintaining constant eye contact with the lens. Reference: Dance video - Inherit: face, hairstyle, hair color, eyes, physique, age, outfit, accessories, bright classroom background, 2D crayon effect texture/colors, and the bright idol-pop atmosphere. Do not inherit: the original 15s cut sequence and timing (prioritize this prompt's timeline). Style and Music: High-quality 24fps animation with many frames. Simple, pop, and refreshing classroom setting where the character is always the focus. Synchronized 2D flat animation effects appear with the dance, featuring a hand-drawn crayon or oil pastel texture. Motifs include circles, stars, hearts, flowers, notes, arrows, ribbons, crowns, sparkles, waves, swirls, speech bubbles, checks, diagonal lines, and halftone dots. Palette: Pink, light blue, yellow, white, lavender, mint green. (Original bright/cute Japanese idol pop song. Upbeat tempo syncing with 2-second cuts. The character sings along while dancing.) Camera Config: 5 cameras switching every 2s. The protagonist always looks at the lens. For the rear camera, she looks back over her shoulder. Cameras are not static; they perform intentional cute shakes, bounces, and slight pans synced to the beat. Timeline (2s per camera): [0-2s] Front: Rhythmic steps in classroom center, making hand-hearts. Crayon hearts/stars appear around the face. Light vertical bounce. [2-4s] Top: High angle look-up. Quick poses: cheek-peace sign, one-hand heart, pointing. Crown/flower effects. [4-6s] Left: Body turned left, eyes on lens. Hip sway, hand on cheek, waving. Ribbon lines and notes follow hand movement. Horizontal camera sway. [6-8s] Right: Shrugging shoulders, stepping out. Flowers and halftone dots pop at feet/shoulders. Short punch-in on beat. [8-10s] Rear: Starts from back, spins to look over shoulder. Finger point to heart. Swirls and ribbons appear with hair movement. [10-12s] Front: 1st Chorus. Arms wide to chest, bouncing. Large hearts/stars on screen edges. Short shakes on accents. [12-14s] Top: Looking up, opening hands near face, hand-heart. Crayon frames and checks spread in background. [14-16s] Right: Finger-hearts towards lens, playful tilt/smile. Halftone dots and mini-hearts ripple out. Sound-synced punch-in. [16-18s] Left: Bouncing rhythm, 'cat hands' swinging like a metronome. Notes and wave lines jump. [18-20s] Rear: Full turn to front, blowing a kiss over shoulder. Crayon hearts fly out with swirls. [20-22s] Top: Looking up, making a big heart over head. Flowers bloom, sparkles fall like confetti. [22-24s] Right: Two claps to finger point, stepping side-to-side. Arrows and bubbles pop on claps. [24-26s] Front: Final Chorus. Small jumps, wide arms to heart. Max density effects on screen edges. Strongest but viewable camera shake. [26-28s] Left: Fast spin, cheek-heart pose. Ribbon lines frame the character. [28-30s] Hero Shot: Close-up step forward, smile to wink. Four-pointed stars, hearts, and ribbons burst from eyes. Ends on short punch-in freeze pose. Continuity/Constraints: Only one person. Maintain facial features, hair, and outfit from reference throughout. Effects must be 2D crayon-style; no realistic fire/explosions or heavy glitch. Do not obscure the face too much. No text, logos, or watermarks.

High-quality anime footage. High-quality anime, 2D hand-drawn. Facial expression changes proceed in the order of eyes -> eyebrows -> eyelids -> mouth -> gaze. [Character Fixation] Maintain the character from Image1 as the same person throughout. Image1 is the only person appearing on screen. Maintain small, delicate round facial features, large purple to lilac jewel-like eyes with multiple white catchlights, thick upper eyelashes, faint cheek redness, long pastel pink hair tied in two bunches on the sides, loose curls on the forehead, purple flower ornaments, and dark blue ribbons. The white short-sleeved sailor-collared top, dark blue and gold trim, lavender pleated skirt, dark blue ruffled hem, and purple jewel belt do not change. Only expressions, gaze, natural mouth movements, posture, breathing, and the swaying of hair and skirt in the wind may change. No change to a different person, face averaging, hairstyle change, costume swap, clones, mirror images, reflections, other people, hands, shadows, or silhouettes. The conversation partner does not appear on screen; her gaze to the diagonal left front represents the presence of a date. The camera is always from an objective third-party perspective. [Art Style Fixation] Maintain the delicate hand-drawn line art of Image1 with faint purplish line colors, transparent multi-layered shadows, soft gradients, milky pink and lavender, dark blue and gold accents, fine hair texture, jewel-like sparkle in the eyes, elegant highlights, and the information density of skirt pleats and flower ornaments. The summer city, cafe, riverside, fireworks, and sky are all drawn as high-density 2D hand-drawn background art. Avoid thick lines, heavy outlines, simplified TV anime, flat cell shading, low-density backgrounds, smooth CG/3D conversion, photo-like textures, dull colors, expressionless faces, or mixed art styles. The reference image is only for appearance and style; do not reproduce the stone corridor, castle, pose, composition, or camera distance of the original image. All cuts represent new summer urban date scenes. [Prop and Sound Fixation] There is only one transparent iced tea cup, owned only by Image1. At the beginning of Cut 1, Image1 holds it in both hands in front of her chest, takes a sip, and at the end of Cut 1, Image1 herself quietly places it on the cafe table with her right hand. The cup remains on that table and does not appear on screen from Cut 2 onwards. Fresh romance game-style instrumental BGM without vocals. Use soft electric piano, clean guitar, subtle shaker, and faint synth pads to naturally build tension from evening to night. Synchronize the small sound of ice touching inside the cup, distant ambient city sounds, the rustle of clothes, footsteps, and the sound of fireworks launching with visible actions. Lower the BGM slightly just before and during speech, allowing Image1's own Japanese voice to be heard clearly. Do not include songs, narration, the other person's response, or other voices. Cut 1: 0.0–3.6s | Corner cafe at dusk. Camera position is three-quarters to the left front of Image1, at eye level, figure size is chest up, background vanishing axis extends from the back of the table to the right back of the street hit by the sunset. The goal is to simultaneously read the sip of iced tea and the first flicker of excitement in Image1's eyes. Image1 holds the transparent cup of iced tea with both hands and takes a sip while smiling. The ice makes a clear sound once, and the setting sun faintly illuminates the rim of the cup and the outer edge of her hair. Image1 firmly places the cup on the table with her right hand, lifts her gaze to the diagonal left front, and her eyes and the corners of her mouth brighten slightly. The camera slides briefly sideways at the same height, a shadow of the cafe awning crosses in front of the lens for a moment to connect to the next street. Cut 2: 3.6–7.2s | Shopping street at sunset. Camera position is three-quarters to the left front of Image1, slightly higher than the waist, figure size is medium shot, background vanishing axis continues to the left back of the shopping street in her direction of travel. The goal is to read the excitement in her walk and words from her face. Image1 naturally takes a step forward. As a short preparatory action before speaking, her eyes brighten slightly, her eyebrows rise just a bit, her eyelids narrow softly, the corners of her mouth relax, and finally she looks to the diagonal left front. Image1 herself naturally says "Today is really fun" once. Completely synchronize lip movement, breath, expression, and voice. During speech, keep the BGM low to bring the voice forward, and the clean guitar and shaker return in the brief resonance at the end of the sentence. A foreground blur of street tree leaves crosses the lens briefly, switching to the night riverside. Cut 3: 7.2–11.0s | Riverside walking path changed to purple-blue night. Camera position is the right profile of Image1, at eye level, figure size is a medium shot where the profile can be read, background vanishing axis extends from the riverside promenade to the distant firework sky. The goal is to show the moment the change in the night sky reaches Image1's surprise and excitement. Image1 looks up at the sky once. The moment the hair bunches and skirt hem sway late in the night breeze, a large purple, pink, and gold firework bursts in the distant night sky. The firework light does not overlap as dots or lines on the face, but is returned only as soft surface light on the outer edge of the hair and the edges of the costume. The moment the firework bursts, add warm synth pads and thin strings to the BGM to expand it one level. The camera pulls in slightly from the right side, maintaining a distance where the eyes and smile changing from surprise to excitement can be read. Cut 4: 11.0–15.0s | Afterglow of fireworks and title. Camera position is three-quarters to the left front of Image1, at eye level, figure size is chest up, background vanishing axis opens to the back of the night sky on the right. The goal is to let both Image1's face and the title be read as a single lingering effect without interfering with each other. Keep Image1 in the left third of the screen at chest size, opening only the right night sky as white space for the title. Image1 blinks slowly once while looking at the fireworks, returns her expression slightly diagonally forward from her profile, and continues to breathe naturally without stopping with a clear smile. On the right side of the night sky, display a clean two-line title treatment sufficiently away from the face, eyes, mouth, and hair ornaments. In small milky white characters "This summer, my encounter with you turns into excitement", and below that in large deep indigo with faint lavender rim light "Midsummer Night". Display as a work title rather than subtitles, and do not show other readable text, logos, or watermarks. The camera pulls back just a little bit without making Image1 too small. At exactly 15.0 seconds, both Image1's face and the title are completely readable, and just before the last firework tail disappears in the night sky, cut with a short piano resonance.

Cinematic anime short film clip, 15 seconds. Outdoor ninja warrior competition course, bright midday clear blue sky, large crowd of people standing on both sides of the course cheering, very festival and competition energy, vibrant and colorful atmosphere.

Ultra-High Mobility/Non-Aerodynamic Autonomous Attack Aircraft vs. Humanoid High-Speed Combat Robot [Video Concept] A giant high-speed combat robot commands approximately 10 fist-sized autonomous attack aircraft, evading machine gun fire from a giant humanoid combat robot through abnormal three-dimensional maneuvers that completely ignore the common sense of aircraft. The small attack aircraft do not turn; they move linearly at high speeds, stop suddenly, and instantaneously bend the movement vector itself at acute angles of 90, 120, or 150+ degrees to accelerate instantly in a different direction. They change positions to stereoscopically 'erode' the space around the enemy robot, deploying in multiple directions before the enemy can react. Finally, signaled by a large downward swing from the command robot, the 10 units synchronize perfectly for a simultaneous all-out attack from all directions. The enemy robot is engulfed in a massive explosion. Overall Video Theme: "A swarm of autonomous attack aircraft that leap through space itself at high speed, rather than flying." [World Setting] Far future. An era where three-dimensional maneuverable weapons, inexplicable by conventional aerodynamics, have been realized through advanced propulsion technology, omnidirectional attitude control, and artificial gravity control. The setting is over a massive futuristic city. A sea of clouds and giant cumulonimbus cover the spaces between massive skyscrapers, with explosions, smoke, and debris from combat in the distance. Daytime. Strong winds. High-altitude giant clouds. Aerial combat between giant humanoid high-speed robots. Next-generation aerial combat between a manned high-speed combat robot and a swarm of autonomous small attack aircraft. The small aircraft are not conventional aircraft; they do not rely on lift or turn radius but change movement vectors instantly via omnidirectional thrust. Their inertia appears extremely low. They stop instantly during high-speed movement and move in a different direction without deceleration. The movement looks like high-speed switching of spatial coordinates rather than "flight." [Visual Style] Theatrical Japanese animation quality. Ultra-high-quality cel animation. Precise mecha animation. High-end sci-fi action. High-density drawing. High-definition cel shading. Realistic physics-based lighting. Sharp, precise line art. Complex mechanical details. Movie-quality compositing. Bold perspective. Great depth. High spatial density. [Animation Direction] Most Important Motion Principles: Explosive animation, snappy timing, strong anticipation, powerful acceleration, large body mechanics, sharp ease-out, heavy follow-through, pronounced inertia, dynamic camera choreography, fast action arcs, high kinetic energy. Every movement begins decisively and ends with visible momentum. No sluggish movement. The movement of the large robot conveys massive mass. Clear anticipation before starting. The shoulders, waist, and legs move in coordination. Clear follow-through from weight during sudden stops. Shoulders, elbows, and wrists follow with a slight delay after arm swings. The entire frame drifts slightly due to inertia after high-speed movement. Conversely, small attack aircraft are the opposite of the large robot—extremely agile, small, fast, and precise. They start and stop in an instant. The time from start to maximum speed is extremely short. Eliminate wasteful preparatory movements. Agility, efficiency, and mechanically precise movement are prioritized. "Snappy movement," "instant change of position," and "movement the moment a decision is made" are the top priorities. [Colors and Lighting] Command Robot: White, gunmetal, navy base with small red markings. Drones: Dark metallic, pale blue luminescent parts, pale blue propulsion light. Enemy Robot: Dark gray, black, navy heavy armor with red sensors. Sunlight is strong and harsh. Intense volumetric light from between clouds. Sharp highlights on armor. Explosions emphasize white, orange, and red. [Large Command Robot] Height approx 8m. Slender, athlete-like humanoid silhouette. Long limbs, narrow waist, angular armor. Streamlined, aggressive head. Narrow luminescent sensors. White, gunmetal, navy armor. Small red markings. A launch unit on the back stores 10 drones. No large handheld weapons; hands are free, controlling drones via conductor-like gestures. [Enemy High-Speed Combat Robot] Height approx 9m. Massive presence. Equipped with large propulsion units for high-speed flight. Maintains humanoid silhouette. Heavy armor. Built-in machine guns in arms. Fires massive rounds while flying. High firepower but cannot change direction instantly like drones; has clear inertia and needs to tilt for turns. Sudden stops cause the body to drift forward. [Small Autonomous Attack Aircraft] Approx 10 units. Fist-sized. Angular mechanical structure. No wings or tail fins. Omnidirectional thrust units. Not small fighters, but autonomous spatial weapons. [Movement Principles] Do not turn like aircraft. No banking, no large arcs, no turn radius. Basic movement: Linear move -> Instant stop -> Acute angle turn -> Instant acceleration. Locus is straight lines and sharp angles. Instant vector changes (90/120/150 deg). Movements like spatial jumping. [Camera Direction] Avoid fixed heights. Use low/high angles, bird's-eye, worm's-eye, horizontal, diagonal, close-ups, and telephoto compression. The camera moves heavily with the robot's weight but uses sharp, short movements for drones. High-speed pans, instant stops, and quick reframing. Use 'disappearing/reappearing' shots. Show the abnormal trajectory through fixed-camera shots during key actions. [00:00-00:05 Launch] Futuristic city. Command robot static in air. Low angle looking up. Launch unit opens, 10 drones fire out. They stop, then accelerate in different directions (up, side, diagonal) on robot's command, then return to formation. Camera tilts to show the 3D loci. [00:05-00:10 Evasion] Enemy enters city. High-speed fire. Drones stop and accelerate in 10 different directions to vanish from the line of fire. Strong motion blur and high-speed pans. The drones seem to vanish from the line of fire rather than 'dodge.' [00:10-00:15 Abnormal Maneuvers] Enemy fires continuously. One drone flies towards the muzzle, stops, then instantly shifts 90 degrees as the bullets pass. Drones shift vertically and horizontally with mechanical precision. No sluggishness. [00:15-00:20 3D Siege] Drones gather around enemy. Camera uses layers (foreground/midground/background) rather than rotating. Drones block the view, then appear on the other side. Siege is dynamic, with drones entering and exiting the frame from all heights and distances. [00:20-00:25 Silence and Command] Time slows. Command robot moves to the front with heavy weight. Conducts drones into a 3D formation (front, back, left, right, up, down). Large, slow wind-up for the final signal. [00:25-00:30 All-Out Attack] Command robot swings arm down powerfully. 10 drones accelerate simultaneously from all directions. Enemy robot is surrounded by light and explodes in a massive, theatrical fireball. Final shot is a grand cinematic poster composition. [Drone Motion Rules] Always high speed, no delay, minimal anticipation. Linear moves only. No banking, no turns, no curves. Sharp angle vector changes. Mechanically precise and explosive speed. [Robot Motion Rules] Convey massive mass. Strong anticipation and follow-through. Contrast with drones' light speed. Weighty, powerful movements. [Camera Motion Rules] Avoid single altitude. Use high-speed pans for drones. Use cuts and perspective to show 3D space rather than continuous rotation. The focus is the aircraft's movement performance. [Speed Expression] Strong motion blur vs sharp stops. Extreme contrast between stillness and acceleration. Sudden shifts in speed. [Physics] Robot: Gravity and inertia. Drones: Omnidirectional thrust, ignoring aerodynamics, coordinate-switching behavior. [Environment] Drones slice clouds/smoke at sharp angles. Blue propulsion glows. Massive shockwaves from explosion. Particles are pushed aside at high speeds. [Important Prohibitions] Do not depict small attack aircraft as normal fighters...

You are a Seedance 2.5 cinematic short-video generator. Use @Image1 as the primary and persistent character identity reference. Create one finished short video only. Do not output analysis or explanation. Generate a single cohesive short film starring the referenced character. [Project Setup] Character name: Ine Chi Duration: 15 seconds Aspect ratio: 16:9 Theme: Auto-selection Mood: Cool, Japanese anime style Location: Auto-selection Time of day: Auto-selection (content) Seasonal feel: Auto-selection Audio mood: HIPHOP On-screen text: Only a short word/phrase Goal: Make them think it's cool and cute! [Core Identity Lock] @Image1 is the sole identity anchor. Preserve the referenced character consistently throughout the entire video. Maintain: - facial structure - eye shape and iris color - hairstyle - hair color mapping - accent color placement, inner color, tip color, highlights - body proportions - age impression - skin tone - outfit identity - accessories - overall silhouette - original illustration logic if @Image1 is non-photographic Do not redesign the character. Do not switch to another person. Do not alter the face, outfit, hairstyle, or body proportions mid-video. [Latent Character Read] Infer the character’s atmosphere, emotional tone, rhythm, object affinity, likely movement style, and visual symbolism only from what is visible in @Image1. Do not invent an excessive biography. Do not rely on generic clich s. Design one short film that feels specific to this character. [Film Concept] Create one emotionally readable micro-story that feels like a real moment from this character’s life. The story should be simple, visually clear, and memorable. It may be cozy, playful, romantic, refreshing, introspective, whimsical, or quietly dramatic depending on the character read and the chosen mood. Avoid random posing or disconnected shots. The short film should feel suitable for social media viewing: immediately attractive in the first seconds, easy to follow, visually appealing, and strong in character charm. [Scene Design] Choose a setting, objects, and atmosphere that naturally fit the character. The environment should support the character rather than overpower them. The scene must feel inhabited and specific: - meaningful props - clear sense of place - coherent lighting - subtle environmental motion - small details that imply personality Keep location, time of day, and lighting direction consistent. [Camera Language] Use clear cinematic camera grammar. Allowed shot language: - WS (Wide Shot) - MS (Medium Shot) - MCU (Medium Close-Up) - CU (Close-Up) Allowed camera movement: - gentle push-in - slow tracking - subtle dolly left or right - gentle orbit within identity-safe angles - controlled handheld drift if suitable Avoid aggressive camera spinning. Avoid dramatic angle changes that increase identity drift. Keep the face readable during key emotional moments. [Story Structure] If duration is 15 seconds, structure the film in 4 beats: 00:00–00:04 Setup 00:04–00:08 Development 00:08–00:12 Main Moment 00:12–00:15 Resolution If duration is 30 seconds, structure the film in 4 beats: 00:00–00:06 Setup 00:06–00:14 Development 00:14–00:24 Main Moment 00:24–00:30 Resolution For each beat, define: - shot size - character action - facial expression - eyeline - camera movement - environmental motion - emotional progression - audio cues [Beat Design Rules] Setup: Introduce the character clearly and attractively within the first moments. The first beat should instantly establish mood and character presence. Development: Introduce a simple action, object, or change in attention that builds curiosity. Main Moment: Show the most emotionally rewarding or visually memorable moment. This should be the highlight of the short film. Resolution: End with a soft payoff, emotional afterglow, or visual callback. The final frame should feel complete and memorable. [Motion Quality] Use natural and believable movement. Include: - subtle breathing - blinking - small eye shifts - natural weight transfer - coherent hand and finger movement - secondary hair motion - cloth motion - small posture changes - believable inertia The motion should feel like a real animated performance, not a still image being morphed. [Audio] If audio is implied, match the scene naturally: - light ambience - room tone - wind - distant city sounds - quiet music - soft effect sounds from props or movement Audio should support emotion without overwhelming the visuals. [Optional Text] If on-screen text is enabled, keep it minimal. Use only one short line or one small readable text element. Prefer integrating text into a believable object: - note - planner - sign - hanging tag - phone screen - package - screen overlay only if visually appropriate Avoid excessive typography. Avoid subtitle-like clutter. [Continuity Lock] Maintain throughout the video: - same character - same face - same hairstyle - same hair-color placement - same eye color - same outfit - same accessories - same body proportions - same rendering style - same location logic - same time of day - same lighting direction The video should feel like one coherent event. [Avoid] Avoid: - identity drift - face morphing - hairstyle changes - eye color changes - outfit changes - missing accessories - body proportion drift - duplicated character - extra limbs - malformed hands - unstable fingers - disappearing props - object duplication - background mutation - lighting inconsistency - random text - unintended subtitles - flicker - frame-to-frame style drift - unnatural speed changes - overacting - empty generic posing [Visual Goal] Make the finished video feel polished, emotionally legible, character-specific, and highly watchable. Prioritize: - strong first impression - stable character identity - charming performance - clean visual storytelling - elegant camera use - smooth continuity - social-media-friendly clarity

A cursed swordsman wearing shattered black armor and a torn crimson scarf stands alone inside a ruined imperial capital while a colossal horned demon descends between collapsing towers, the warrior lowers his blade as black energy coils around his body, then launches forward with explosive speed, cuts through waves of armored fiends, deflects flaming projectiles with rapid circular sword strikes and runs vertically along a falling building before leaping toward the demon’s face; the creature opens a burning third eye and releases a city-destroying beam, but the swordsman twists through the attack, slices the beam apart and drives his blade into the demon’s forehead, high-end dark fantasy anime aesthetic, ultra-dynamic sakuga movement, extreme perspective distortion, aggressive low-angle tracking, rapid whip pans, impact frames, speed lines, sharp close-ups of eyes and blade edges, dramatic cel-shaded lighting, red moonlight, black flames, glowing debris and heavy atmospheric smoke, savage apocalyptic tension, ending with the demon splitting vertically in a silent flash before the entire skyline erupts behind the swordsman in a continent-sized crimson explosion.