An Anime Character Breaks through Giant Negative Type

A character passes black Japanese typographic obstacles before a red message of encouragement appears.

Generated results

Mode not specified

Prompt metadata
Shared by	@LilyLoveSNPY
Model	MiniMax H3
Category	Animation
Dataset	SkyNotSilent · awesome-MiniMax-H3-cases

The prompt

CHARACTER (identical in every cut): a woman in her twenties, short black hair, grey blazer over a white shirt, red high-top sneakers. Same face, eyes and proportions in all four cuts - do NOT redesign her between cuts; exaggeration applies to expression and pose only, never to face structure. She moves with fluid athletic grace, never stiff.

WORLD: an endless white void with a pale-grey grid floor, dust motes in hard rim light. Stylized 2.5D anime aesthetic, photorealistic high-detail 2.5D CG, clean cel-shaded face, high-key white and grey with one red accent, soft bloom, lens flares, 35mm grain.

TEXT (critical): exactly three pieces of text appear, one at a time, never together. Each is a single horizontal line spanning the full width of the frame, razor sharp, correctly spelled, clean spacing between characters.
(1) Japanese, nine characters: 失敗を恐れすぎだよ - thick wet black brush ink.
(2) Japanese, thirteen characters: 人の目ばかり気にしてるよね - thick wet black brush ink.
(3) English, four words: You can do it! - bold, solid red.
No other letters, numbers, signage, captions or subtitles anywhere; the spoken lines must never appear as text.

SHOT 1 (0:00-0:04) [Truck right]. Emotion: irritated defiance.
0:00-0:01: extreme close-up of a red sneaker landing on the white floor, black ink cracking outward.
0:01-0:04: camera drops to ground level and trucks right alongside her as she walks - text (1) erupts from her footprints and locks into the air behind her, taller than she is. She says in Japanese: "うるさいな" - flat, annoyed, half over her shoulder, lip-synced.

Cut to SHOT 2 (0:04-0:08) [Tracking shot]. Emotion: pressure.
0:04-0:06: she breaks into a run, three hard strides, foreground grid slabs whipping past in motion blur, speed lines streaking.
0:06-0:08: text (1) dissolves and text (2) rears up as a long ink wall chasing her, dripping, slamming down behind her heels twice and cratering the floor. Camera tracks low behind her shoulder, shaking on each impact.

Cut to SHOT 3 (0:08-0:12) [Push in]. Emotion: sharp refusal.
0:08-0:09: she skids to a dead stop, one arm swinging out, dust bursting sideways.
0:09-0:10: hard accelerating push-in as she pivots and shoves both palms into the falling ink wall - a single-frame high-contrast inverted impact frame, shockwave ring.
0:10-0:12: text (2) shatters into black ink fragments scattering past camera in slow motion.

Cut to SHOT 4 (0:12-0:15) [Pull out]. Emotion: light, decided.
0:12-0:13: she steps through the fragments and stamps a foot down.
0:13-0:15: text (3) rams up out of the floor beside her, filling the frame, as the camera pulls out to a wide of the void. She turns her head and gazes directly into the camera. She says in Japanese: "やってみる" - light, half smiling, lip-synced.

CAMERA GRAMMAR: one move per cut; shot sizes swing hard - ECU, low tracking wide, ECU, wide. PERMITTED: dolly, push-in, truck, tilt, low tracking. AVOID: whip pans, handheld jitter, 360-degree rotation, slow zooms.

AUDIO: sneaker squeak, running footfalls, a heavy wet ink slam on each impact, a glass-like shatter on the shove. Music: minimal electronic percussion, one beat drop on the shatter, one on the final text. Only one voice in the clip - hers. No narration, no second speaker.

CONSTRAINTS: she is the only person in frame. Stable geometry, natural motion, no warping.

Try in DraPub

More prompts from the same visual neighborhood.

Back to prompt gallery

Animation

A Steam Train Passes a Low-Poly Castle

A black steam locomotive pulls back from a close-up of its working wheels, hauling red carriages across an arched bridge as a medieval castle and river enter the wide view. Native-video output: 15s · 480×480. The creator published the complete prompt, preserved here verbatim.

Animation

A Golden Celebration Sphere Unfolds into Thank-You Typography

Golden sparks introduce a celebration sphere, fireworks, and confetti. The sphere opens to reveal a 5,300-follower milestone, the creator's name, and thank-you lettering. Native-video output: 15s · 2560×1440. The creator published the complete prompt, preserved here verbatim.

Animation

A Cat-Eared Chef Finishes a Plate of Fried Rice

A blue-haired cat-eared chef stir-fries rice in a Chinese restaurant, works through flame and exertion, plates the dish, and serves it to a customer. Native-video output: 15s · 480×800. The creator published the complete prompt, preserved here verbatim.

Animation

A Drawing App Builds a Red-Haired Character from Blank Canvas

A drawing interface moves from blank canvas through sketching, color, and shading to complete a full-body red-haired character design. Native-video output: 15s · 2560×1440. The creator published the complete prompt, preserved here verbatim.