Won't Share a Bite of Ramen · Anime Couple Mini Skit

PixVerse V5.52D AnimationAnimationComedyRomanceFoodStructured Prompt5s
Japanese anime look with the performance written second by second: she leans in for a taste, he shields the bowl and turns away, and the whole mood is carried by body language.

PROMPT · Video Prompt

Create with AI
VIDEO PROMPT (Japanese anime look, 16:9) SETTING: The dining nook of a snug Korean apartment. On the table sits a bowl of ramen in a fiery red broth. Warm indoor light; crisp anime linework with smooth cel-style shading. CAST: 1) A Korean man in his twenties: thick black-framed glasses, plain white T-shirt, chopsticks in hand. 2) A Korean woman in her twenties: hair in a ponytail, sky-blue T-shirt. BEATS (about 5 to 7 seconds in total): - 0.0–1.5s: She bends toward the bowl and motions at it, either pointing or making a small reach, to ask for a taste. - 1.5–3.0s: Visibly annoyed, he drags the bowl a little nearer to himself, his shoulders stiffening. - 3.0–5.5s: He swings his body away so his back faces her, an unmistakable "hands off" stance, and carries on eating with his chopsticks, faster and more stubbornly than before. She stays half-frozen on the spot, irritated yet still hoping for a mouthful. LINES (Korean, spoken with natural pronunciation): - Woman (teasing, persistent): “한 입만!” (Romanization: "Han ip-man!") - Man (cross and louder): “안 돼!” (Romanization: "An dwae!") SOUND: - Quiet room ambience and no music. - A faint slurp or chopstick clatter is welcome at low volume. - Voices should be clear and clean, acted like ordinary life at home. CAMERA: - One medium shot holding both people and the bowl inside the frame. - No fast cuts; movement stays smooth and easy to follow. AVOID: Nudity, gore, on-screen text or watermarks, extra fingers, warped faces, brand logos.

✍️ Editor’s Notes

The skeleton here is a sectioned checklist: SCENE, CHARACTERS, ACTION / TIMING, DIALOGUE, AUDIO NOTES, CAMERA and NEGATIVE each get their own block, so the model always knows where to look for each kind of information. The most valuable part is ACTION / TIMING, which splits a little over five seconds into three beats, each carrying just one action and one emotional shift: leaning in to ask, guarding the bowl, turning away in a huff. The body language is concrete (tense shoulders, body angled away, eating faster and more stubbornly) instead of piling up emotion adjectives. The dialogue is given in the original Korean plus romanization, which anchors both pronunciation and mouth shapes. CAMERA locks a medium shot with both people and the bowl in frame, then adds a no-rapid-cuts line so the short doesn't get chopped up. NEGATIVE sits last and lists only generic risks. The reusable trick is the timing format of one action and one emotion per beat.

Create with AI

Related Works

׋›
↑