Anime Katana Duel Against Walker Droid in Ruined Hangar

A 2D anime sakuga fight sequence where a swordswoman in a teal jacket sprints through a flooded orbital hangar and destroys a quadruped walker droid with a katana. The prompt uses a storyboard reference for shot order plus two character reference sheets, and specifies beat-by-beat staging, impact frames, and a silent hitstop on the final head strike.

This clip could not be loaded — it may have been removed from X.
Alessia @alessia_orbit ·
View on X ↗

Prompt

Steel Match

She wasn't built for this fight. Neither was it.

ChatGPT + Seedance 2.0 via Dreamina
Prompt + character sheet 👇🏻

Use [Ref storyboard] only for shot order, camera, staging, screen direction, and spatial continuity. Render ONLY the beats listed below; skip storyboard panels not listed. Render from these beats and mapped references, not the sheet itself; ignore malformed anatomy, wrong props, extra limbs, stray marks, arrows, icons, borders, typography. Read each beat as a separate shot; preserve identities, never merge poses. Use [Ref 1] as sole authority for C1 face, body, dark hair bun, knotted off-shoulder teal jacket, loose printed pants, sneakers, katana design, proportions, likeness; use [Ref 2] as sole authority for C2 quadruped walker droid hull, stepped superstructure, leg set, cannons, sensor visor, proportions, armor. C2's twin cannons are its sole weapon origins.

VISUAL STYLE: 2D anime sakuga, cel shading, hard sun through a broken hangar roof. Cool figure against cold wet-steel hangar; C2's amber optics and C1's glowing cyan-teal blade are the opposing light accents in frame. Impact frames, sharp inserts, dust haze, speed smears, plasma sparks.

AUDIO: No score. Servo chatter, blaster pulses, blade ring, fast steps, shield buzz, water hiss, breath cuts. Total silence for one beat on the P14 held impact, then sound crashes back.

ENVIRONMENT: Shallow mirror-water basin in a ruined orbital hangar, long central lane, broken roof ribs, two pylon blocks, open sky glare.

EMOTIONAL GUIDANCE: C1 is already sprinting at open; every pause is a breath between hits; the last image is a clean upright finish.

RHYTHM + ESCALATION: Begin on impact. Fast serial cuts through the barrage, a quick held tap on the leg ricochet, spike through the vertical climb and drop, full hitstop on the head strike as the single held impact of the film, then release into collapse.

BEATS:

P01: Action master: C2 crab-steps while firing down the wet lane as C1 sprints in from screen left, katana raised, roof ribs and pylons locking geography.

P02: Low water catch: C1 runs through ankle-deep water, catches one bolt on the sword edge as a second streaks wide, spray thrown backward in a hard diagonal.

P04: Leg ricochet: C1's blade arc redirects a skipped bolt into C2's front leg joint - brief held frame on the hit, water hissing - first clean score.

P09: Long-lens crush: all incoming fire comes from C2's side only; C1 darts through a cyan blade cage, shoulders low, feet skipping over wet impacts.

P11: Hull run-up: C1 charges up the droid's moving front leg and hull edge, body low, katana tucked tight as cannon fire rakes beneath her.

P12: Vertical switch and drop: low upward angle from C2's head as C1 clears the hull, turns the katana vertical point-down, and falls straight toward the exact head center, a vertical streak against the roof glare.

P14: Head impact: the katana punches into C2's head center - held impact frame, one beat of frozen silence, water droplets hanging mid-air - then rips free in the same vertical line as the hull snaps backward.

P15: Moving collapse: high wide release as C2 crashes screen right with a split head plate, while C1 lands upright nearby holding the katana.
? How to use this prompt ▾
  1. Copy the prompt using the green "Copy prompt" button above.
  2. Open your model — launch Seedance 2.0 on its official interface.
  3. Paste & generate — keep the verbatim prompt text for best adherence. Set aspect ratio to 9:16 and clip duration to 15s.

What makes this prompt work

Every insight below is verified against benchmark generation parameters, not asserted. How our prompt engineering benchmarks work →

CAMERA
"low-angle" is a core cinematic camera technique for establishing spatial immersion and subject framing. Also present: "wide shot", "tracking", "slow motion".
LIGHTING
"backlight, rim light" establishes luminance depth, volumetric shadow falloff, and tonal contrast across the generation.
MOOD
"energetic, dramatic" — Mood and aesthetic tone are explicitly directed to maintain emotional consistency throughout the scene.
DURATION
At 15s, this clip is optimized for high temporal coherence and fluid, realistic motion physics.
MODEL
Seedance 2.0 (ByteDance) provides the underlying diffusion and motion architecture. Camera, lighting and duration cues transfer across models — what differs is how literally each model interprets motion.
PROMPT LANGUAGE
The prompt is written in English. This library keeps every prompt in its original language rather than translating it — the exact words are what was conditioned during generation.
Seedance 2.5 BytePlus Cloud Partner

Tell bigger stories with Seedance 2.5

Create AI videos up to 30 seconds long with richer image, video, and audio references. Ideal for complex ideas that need consistency and control.

Explore on BytePlus Official ByteDance Enterprise Cloud