Dance creators
Drop a fresh routine without booking a studio. The AI street dance template keeps the same dancer across drops.
One character,
one high-energy street dance.
Edition 07 β A two-step street dance
Step one writes a character design sheet β four full-body dance poses plus a three-angle head strip of a single 18-year-old Asian street dancer on a gritty industrial backdrop. Step two hands the sheet to Seedance i2v and renders a ten-second cinematic hip-hop dance inside a sun-lit warehouse with low-angle orbit camera and dust-lit golden sunbeams. The whole thing happens in the console below.


The prompts below are the verbatim source. Copy them into the console above and tweak the dancer, location, or wardrobe.

A character-design brief for GPT Image 2.0. Prioritizes face fidelity through a multi-angle head reference strip (front, side, 3/4) while four full-body action poses lock the dancer's silhouette, wardrobe, and body proportions.
Character design reference sheet, focused on a single 18-year-old Asian female street dancer. Includes 4 large, highly-detailed full-body dynamic dance poses (highlighting dance movements, clear faces). To the side, a clear multi-angle reference strip containing only 3 detailed head close-ups (front, side, 3/4 angle). Minimize text elements, prioritize pixel space for facial detail rendering. Background: gritty industrial style with photorealistic lighting and shadows.
CHARACTER LOCK:
Long wavy black hair, athletic build, white short-sleeve tank top, loose beige cargo pants, expressive young face. Same dancer across all 4 full-body poses and the 3 head close-ups.
LAYOUT:
4 full-body poses arranged in a 2x2 main grid (left/center area).
3 head close-ups arranged as a vertical reference strip (right side).
No panel borders, no captions, no numbers, no watermark.
STYLE:
Photoreal, high-detail, sharp focus on faces, cinematic warm-cool color grade, gritty industrial backdrop with concrete, exposed beams, and warm directional light.Model β gpt-image-2
A Seedance 2.0 i2v shot list. The character sheet locks the dancer; the prompt scripts the warehouse environment, golden sunbeam dust, low-angle orbit tracking, and the expressive close-up payoff on the dancer's face.
Use the provided character sheet as reference.
CONCEPT:
Cinematic street dance β 10 seconds of high-energy hip-hop movement inside a sun-lit industrial warehouse.
CHARACTER LOCK:
Same 18-year-old Asian woman from the reference sheet β long wavy black hair, white short-sleeve tank top, loose beige cargo pants. Do not duplicate her, do not drift in identity. Face fidelity is critical.
SCENE:
Large industrial warehouse, dust motes catching in golden sunbeams from high windows. Polished concrete floor, exposed beams, warm late-afternoon light. No clutter, no extras.
ACTION:
High-energy hip-hop choreography β powerful dynamic moves, hair flying in the air, sweat glistening on skin, expressive facial micro-emotions. Camera focused on her expressive face during key beats.
CAMERA:
Low-angle orbit tracking around the dancer. Smooth circular dolly. Occasional cuts to close-up on her face during accents. Subtle handheld realism, shallow depth of field.
STYLE:
4K, ultra-realistic, cinematic, masterpiece quality. Dust dancing in the golden sunbeams. Sharp focus on the dancer, soft fall-off on the warehouse background.
RESTRICTIONS:
No text overlays, no captions, no watermark, no slow-motion gimmicks, no anime stylization, no identity drift across cuts.Model β seedance-2-beta
Once you've tuned a dancer you like, save the prompt pair and re-run for every track, genre, or location.
Browse all templates
The AI street dance template is a two-pass workflow that produces a 10-second cinematic hip-hop street dance from text alone. The first pass writes a character design sheet with 4 full-body dance poses and a 3-angle head reference strip. The second pass animates the dancer into a high-energy warehouse routine with cinematic camera motion.
You run the AI street dance template inside the console above. Step one calls GPT Image 2.0 with a character-design brief that prioritizes face fidelity. Step two hands the sheet to Seedance 2.0 image-to-video. Swap the dancer, swap the location, hit generate, ship the AI street dance template's final clip. GPT Image 2.0 image studio Β· Seedance 2.0 image-to-video
Multi-angle head strip
Front + side + 3/4 head reference for maximum face fidelity across the AI street dance template's video pass.
4 action poses
Locks silhouette, wardrobe, and body language so the dancer reads as one person across the routine.
Cinematic warehouse
Sun-lit industrial space with dust caught in golden beams. No green-screen feel.
Low-angle orbit
Camera circles the dancer at low angle for the cinematic hip-hop look that reads on social.
Anyone shipping short-form dance or movement content where the dancer's face and body need to stay consistent on camera.
Drop a fresh routine without booking a studio. The AI street dance template keeps the same dancer across drops.
Build tie-in dance clips for hip-hop releases. The AI street dance template handles the performer and the location.
Showcase a capsule mid-movement. The AI street dance template puts the brand pieces on a consistent persona.
Previs a routine before teaching it. The AI street dance template doubles as character + camera mood board.
Why the AI street dance template beats a warehouse rental for short-form social dance content where consistency matters more than perfect choreography.
See credit pricing Β· Seedance 2.0 Fast Β· Dance tutorial template
The AI street dance template is a paired prompt: a GPT Image 2.0 brief that produces a character design sheet of one street dancer (4 full-body poses + 3 head angles) on a gritty industrial backdrop, and a Seedance 2.0 i2v shot list that animates that sheet into a 10-second cinematic hip-hop dance inside a sun-lit warehouse. Both prompts are on this page and free to remix.
Multi-angle head references give the AI street dance template's step-two video pass enough face data to lock identity across cuts. Without it, the dancer can drift between profiles. With it, the face stays consistent.
Yes. Edit the character description in step one β age, ethnicity, wardrobe, hair. The AI street dance template propagates the lock into step two so the new dancer holds across the routine.
About four minutes end-to-end on Seedance 2.0 Fast. Step one (the character sheet) takes around a minute. Step two (the 10-second video) takes around three.
16:9 landscape β drops cleanly into YouTube, dance compilations, and cinema-aspect social cuts. A safe vertical center-crop works for Reels and TikTok.
Yes. Swap the environment description in both step-one and step-two prompts β rooftop, alley, dance studio, gym. The AI street dance template's character lock holds regardless of location.
Generations from seedance2.so are yours to use. The AI street dance template's prompt is original β just avoid trademarked brand logos on wardrobe pieces.
A plain text-to-video dance prompt usually drifts mid-clip β the dancer's face changes between cuts. The AI street dance template builds a multi-angle character sheet first, so the video pass renders against a locked identity. That is why the dancer reads as one person across the full routine.
Two prompts, one console, one render. Character sheet, then video β or browse all AI workflow templates.