How to Write a Hero-Point Dance Prompt for Looks Can Kill AI

IVE’s “Looks Can Kill” gives prompt writers a very specific visual target: a girl-hero concept with finger-gun and point choreography. IVE is a K-pop girl group under Starship Entertainment. “Looks Can Kill” is the title track of their fifth EP; the single was released on 19 October 2026 and the EP follows on 26 October 2026. At Looks Can Kill AI (lookscankillai.pro), that release is a useful reference point for style and motion language, not official material.

This site is an independent photo/text-to-video studio built on Seedance 2.0. It is not affiliated with IVE or Starship. It does not include the official recording, official video or member likenesses. You may only upload your own photo or a photo you have permission to use.

A hero-point dance clip lives or dies on one readable action. If the prompt tries to describe a full routine, the model often produces soft hands, drifting faces, extra people and camera moves that hide the point. The fix is a tighter prompt structure: subject lock, scene, one action, camera and exclusions.

1 · Upload your photoClick to upload a photoDrag and drop, or click to choose · JPG, PNG or WebP · up to 10 MBOne person, full body, face and shoes visible, even lighting. A clear solo photo keeps the face and outfit stable through the turn.

2 · Describe or pick a sceneUse the person in the uploaded photo as the only hero. Keep their face, hair, skin tone and outfit exactly as photographed. Night rooftop above a city, wind moving their hair and jacket. They step forward, cross their arms, snap one finger-gun hand toward the camera, point at the lens, shift their weight through a sharp side step, raise one arm overhead and hold a final action-hero pose. One continuous vertical 9:16 shot at full-body distance, confident slow motion on the snap, cinematic teal and amber light. No text, no watermark, no other people, no logos.

Example loaded — edit it or generate.

Rooftop heroPractice roomComic frameSpotlight finale

3 · Choose format and length

Resolution

480p720p1080p

Length

5s8s10s12s

Format

9:1616:91:1

Sign in to create my videoThis video: 35 credits · 720p · 12s

This clip: 35 credits · 10 clips a month for $39.90 (Pro), or 102 a year for $319

Why the hero-point dance needs a tighter prompt

The “Looks Can Kill” girl-hero look is built on confidence, clean silhouettes and a direct finger-gun / point gesture. That does not mean the prompt should be vague. A strong clip needs the model to preserve the uploaded subject while performing one clear gesture in a controlled frame. The feature page explains the studio’s dance-video workflow; this guide focuses on the prompt itself.

Short vertical clips are especially unforgiving. A 5–12 second output has no room for three costume changes, a crowd and a full camera orbit. Treat the prompt like a shot list for one hero moment. If you want context on the release and the visual concept, read what “Looks Can Kill” is. If you want the full production workflow, see how to make a Looks Can Kill AI dance video.

The five-part prompt formula

Use this numbered structure every time. It keeps the model focused and makes debugging easier when a render misses.

1. Subject lock: Name the uploaded photo as the only identity reference. Ask for the same face, hair, skin tone and outfit. Say “no other people” and “do not invent a new identity.”
2. Scene: Describe one location, one light setup and one background. Keep it simple: rooftop, studio, alley, stage or desert. Use 9:16 vertical framing by default.
3. One action: Choose a single hero-point action: step-and-point, finger-gun to camera, hip-shift point, over-shoulder point or final pose hold. Do not stack a spin, jump and hair flip in the same prompt.
4. Camera: Specify shot size, angle, movement and duration. “Medium shot, slight low angle, slow push-in, no cuts, 12 seconds” is more useful than “cinematic.”
5. Exclusions: List what must not appear: extra people, text, logos, watermarks, distorted hands, duplicate limbs, outfit changes, camera cuts and fast shake.

What to put in each slot

| Prompt part | Write this | Avoid this | | ------------ | --------------------------------------------------------------------------------------- | ------------------------------------------------------------ | | Subject lock | “Use the uploaded photo as the only subject reference; preserve face, hair and outfit.” | “Make them look like a K-pop star” without identity control. | | Scene | One location, one light source, clean background. | Crowds, multiple sets or busy signage. | | One action | One point, one weight shift, one hold. | Full choreography, spins, jumps and pose changes. | | Camera | Shot size, angle, movement, no cuts, duration. | “Dynamic camera” with no direction. | | Exclusions | No extra people, no text, no watermark, no distorted hands. | Leaving hands and identity to chance. |

Five complete example prompts

Each prompt below is complete. Replace the uploaded-photo instruction with your own photo or a photo you have permission to use. Do not upload member likenesses or images you do not have rights to use.

Example 1: Rooftop hero point

Use the uploaded photo as the only subject lock: preserve the person’s face, hair, skin tone and outfit exactly; do not invent a new identity. Scene: a clean city rooftop at blue hour, soft violet rim light, simple skyline, no logos or signs. One action: the subject steps forward, raises one hand into a crisp finger-gun point toward the camera and holds the final pose. Camera: 9:16 vertical, medium shot, slight low angle, slow push-in, no cuts, 12 seconds. Exclusions: no extra people, no text, no watermark, no brand marks, no distorted hands, no duplicate limbs, no camera shake, no outfit changes.

Example 2: Studio hero-point with a clean pop

Use the uploaded photo as the only subject reference and keep the same face, hair and outfit throughout. Scene: a minimal dance studio with a dark seamless backdrop, one soft key light and a subtle floor reflection. One action: the subject shifts weight to one hip, points a finger-gun at the lens, then freezes in a confident hero pose. Camera: 9:16 vertical, waist-up medium shot, eye-level, locked frame with a tiny push-in, no cuts, 10 seconds. Exclusions: no other dancers, no mirrors with reflections, no text, no logo, no watermark, no extra fingers, no melting face, no sudden zoom.

Example 3: Neon alley point

Use the uploaded photo as the only identity lock: preserve the face, hair, skin tone and clothing; do not change the person’s age or features. Scene: a narrow neon-lit alley with pink and cyan practical lights, wet ground, clean background without readable signs. One action: the subject turns from profile to camera and delivers one sharp point with the index finger, holding the stare. Camera: 9:16 vertical, medium close-up, slight low angle, slow lateral drift, no cuts, 8 seconds. Exclusions: no crowd, no text, no logos, no watermark, no distorted hand, no duplicate arm, no flashing lights, no costume change, no camera roll.

Example 4: Desert girl-hero point

Use the uploaded photo as the only subject reference; keep the same face, hair and outfit exactly. Scene: a sunlit desert plain with a clean horizon, warm late-afternoon light and a simple sky. One action: the subject plants one foot forward and raises a steady finger-gun point toward the camera, then holds the pose. Camera: 9:16 vertical, full-body to medium shot, low angle, slow push-in, no cuts, 12 seconds. Exclusions: no extra people, no animals, no text, no watermark, no brand marks, no distorted hands, no floating limbs, no outfit change, no fast whip pan.

Example 5: Stage finale hold

Use the uploaded photo as the only subject lock: preserve identity, hair, skin tone and outfit; no new face. Scene: a dark stage with a single white spotlight, faint haze and a clean floor. One action: the subject steps into the spotlight and points directly at the viewer with one finger-gun, ending in a still hero pose. Camera: 9:16 vertical, medium shot, slightly below eye level, smooth 15-degree arc, no cuts, 12 seconds. Exclusions: no audience, no text, no logo, no watermark, no backup dancers, no distorted fingers, no duplicate body, no strobe, no outfit swap.

Production notes before you render

On the studio, output is 9:16 vertical by default and runs 5–12 seconds. A standard 720p 12-second clip costs 35 credits; 480p 12s is 20 credits; 1080p 12s is 90 credits. Credits start from $9.90. Failed renders are refunded. One-time packs never expire; monthly credits expire at the end of the billing month; yearly credits are valid for the billing year. Check pricing before you batch-render.

Keep your first pass at 480p or 720p to test the pose and hands. Once the point reads cleanly, move to 1080p for the final clip. If the model adds a second person, shorten the scene description and repeat “no extra people” in exclusions. If the hand warps, simplify the action to one point and avoid describing jewelry or complex sleeves.

FAQ

Is Looks Can Kill AI affiliated with IVE or Starship?

No. Looks Can Kill AI is an independent photo/text-to-video studio built on Seedance 2.0. It is not affiliated with IVE or Starship and does not include the official recording, official video or member likenesses.

Can I upload a photo of an IVE member?

You may only upload your own photo or a photo you have permission to use. The studio does not provide member likenesses, and this guide is about writing original prompts for your own permitted images.

What does “subject lock” mean in a dance prompt?

Subject lock is the instruction that tells the model which photo controls identity. It should say to preserve the face, hair, skin tone and outfit, and to avoid inventing a new person. It is the first part of the five-part formula.

How long should the hero-point clip be?

The studio supports 5–12 seconds with 9:16 vertical output by default. A 12-second clip is a good standard for a full step-in, point and hold, but 8–10 seconds can be enough for a close-up point.

What should I put in the exclusions section?

List the failures you want to prevent: extra people, text, logos, watermarks, distorted hands, duplicate limbs, camera cuts, outfit changes and fast shake. Keep exclusions specific to the shot rather than adding a long list of unrelated negatives.

Do failed renders cost credits?

Failed renders are refunded. Pricing starts from $9.90, a standard 720p 12-second clip costs 35 credits, 480p 12s is 20 credits and 1080p 12s is 90 credits. One-time packs never expire; monthly credits expire at the end of the billing month; yearly credits are valid for the billing year.