COPYABLE VIDEO PROMPTS & CAMERA TIPS

Charme Dance AI Prompt Guide

Learn how to describe the lead role, reference choreography, group scene, identity constraints, camera movement, and stage lighting.

Create Your Video

Upload one photo. The reference video guides the movement. No editing required.

Browse Files

JPG, PNG, or WebP up to 15MB. Best results with a clear front-facing portrait.

Aspect Ratio
Video Quality
Duration

By clicking Generate Video, you confirm that you own this photo or have permission from the adult pictured, and agree to our Terms of Service and Privacy Policy.

Powered by Wan 3.0 via KIE AI. Uploaded images are subject to content safety checks.

From $4.99/videoSee Pricing →

Reference Dance Preview

Original Reference
AI Generated

Upload one photo to create your own Charme Dance AI video. The original 15-second choreography guides the animation with selectable duration and quality.

PUBLIC TEACHING PROMPTS • NOT PRODUCTION SETTINGS

Build a Prompt Around Role, Motion, and Identity

Effective dance-video prompts are specific about who changes, what motion should be followed, and which parts of the scene must stay stable. The examples below are educational prompts only. They do not expose or copy the private production configuration used by the site generator.

Section 01

What Is a Charme Dance AI Prompt?

A prompt is a set of production instructions for a video model. In a dance-reference workflow, it should explain which person is replaced, what the source motion controls, and which visual properties should remain stable. The photo supplies identity, and the reference supplies choreography. The prompt helps the model understand how those inputs relate.

This site’s normal generator does not require you to enter a prompt. Its reference and production instructions are already configured. The examples here are useful for learning how AI dance prompts are structured, testing other tools, or planning a personal project.

Do not treat teaching prompts as internal settings

Public examples are deliberately written for explanation. They do not reproduce the private KIE_WAN_INPUT_JSON or any operational secrets.
Section 02

Anatomy of an Effective Dance Video Prompt

Character role

Name the replacement target: the central black-hat lead, not the background ensemble.

Motion reference

State that the reference controls timing, full-body choreography, group blocking, and camera rhythm.

Identity preservation

Ask for stable face, hair, skin tone, clothing silhouette, and body proportions through motion.

Scene continuity

Preserve the stage, warm lighting, group size, formation, and fire or light accents.

Section 03

Copyable Prompt Examples

Stage-faithful lead replacement

Teaching Prompt

Use the uploaded adult portrait as the identity of the central lead dancer in a 15-second group dance video. Preserve the reference clip’s black hat, full-body choreography, ensemble formation, warm amber stage lighting, camera movement, and fire accents. Keep the face, hair, and body proportions consistent. Do not replace the supporting dancers or redesign the stage.

Avoid

extra lead dancers, changing group size, warped hands, face melting, random camera cuts, text overlays

Recommended: 9:16, 720P or 1080P, reference video required

Identity consistency for turns

Teaching Prompt

Animate the photographed adult as the only central lead dancer. Maintain the same face, hairline, hair length, skin tone, and outfit silhouette through profile turns, arm swings, and partial occlusion by the group. Follow the source timing and stage blocking exactly. Keep the background dancers unchanged and preserve a natural live-performance look.

Avoid

identity swap, changing hairstyle, rubber limbs, duplicated face, sudden close-up, background collapse

Camera and lighting continuity

Teaching Prompt

Create a 15-second full-body dance clip using the provided portrait as the lead identity and the provided video as the motion reference. Maintain the reference’s camera push, ensemble rhythm, warm stage lights, dark background, and brief fire accents. Use realistic fabric motion and stable hands. Keep the lead readable whenever other dancers cross the frame.

Avoid

neon color shift, daylight scene, flickering lights, incoherent stage, excessive motion blur

Square social variation

Teaching Prompt

Adapt the reference dance into a 1:1 social video. Keep the same choreography and group formation while framing the black-hat lead as the identity replacement. Preserve the lead’s face, hair, and body proportions, and retain the original stage atmosphere. Use clean edges and a centered composition suitable for a square post.

Avoid

cropped face, missing feet, stretched frame, blurry ensemble, extra people

Recommended: 1:1, 720P, social caption-safe framing

Section 04

How to Specify the Correct Lead Dancer

The lead is the person the viewer should recognize after the transformation. In this reference, the black hat and central placement make that role visually distinct. A prompt should say “central black-hat lead” instead of “the person in the video,” which could be interpreted as every dancer in the group.

The remaining performers are part of the scene. Replacing all of them with one face creates a duplicated character and weakens the group-performance idea. Ask explicitly to keep the background dancers, formation, and spacing unchanged.

Section 05

Camera Motion, Rhythm & Stage Lighting

Camera language matters because a dance clip can fail even when the face looks right. Specify that the camera should keep the lead readable during group crossings and avoid random cuts that break the reference timing.

Lighting is part of the identity of the scene. The reference uses warm stage tones, darker surroundings, and fire-like accents. If the prompt asks for a completely different color palette, the output may look like a different event rather than a personalized version of the same dance.

Section 06

Common Problems and Negative Prompt Tips

Negative instructions should target visible failures rather than vague quality words. Useful items include warped hands, duplicated faces, changing group size, missing lead, random camera cuts, unstable clothing, rubber limbs, and text artifacts.

If the face changes during a turn, add an identity clause and simplify the scene. If the model replaces the whole group, repeat that only the central lead should change. If the camera drifts, describe the reference camera as the authority.

Section 07

Using Photo and Video References Together

A photo says, “this is the person.” A video says, “this is how the body, group, camera, and timing should move.” Combining them is different from text-only generation because the model receives an explicit motion source instead of guessing choreography from a sentence.

Use a portrait with the same gender presentation, clothing silhouette, or visual proportion you want in the result when possible. Conflicting cues can force the model to invent a compromise, especially around hair, shoulders, and outfit shape.

Section 08

Wan, Kling & Other AI Video Tools

Wan 3.0, Kling, and other video systems expose different strengths and controls. Some favor prompt-driven camera motion, while others are designed around image identity plus a reference clip. Prompt language should follow the actual capabilities of the tool you are using.

The website generator is already configured for its production workflow. The prompt examples here are for comparison, learning, or experimentation—not instructions to paste into the site. Continue with the site workflow or compare results in the example gallery.

FAQ

Charme Dance AI FAQ

See the Prompt Ideas in Action

Use the shared generator to create a reference-guided dance video without writing a prompt.

Related Guides

TRANSPARENT PRICING

Simple Credit Packages for Every Creator

Choose a credit pack to create 15-second Charme Dance AI videos in 480P, 720P, or 1080P and 9:16, 16:9, or 1:1 format.

Credit Consumption: 5s: 10 Credits (480P) · 15s: 15 Credits (480P) · 30 Credits (720P) · 60 Credits (1080P) · 9:16, 16:9, 1:1

Credits are automatically refunded for confirmed failed generations.

Starter

10 Credits
$4.99

Perfect for trying your first 5-second Charme Dance AI video.

Up to 1 Video5 Seconds Only
  • 10 Credits
  • Up to 1 Video (5s · 480P)
  • 5 Seconds Only
  • AI Video from 1 Photo
  • Reference-guided Choreography
  • Standard Preview Quality (480P)
  • No Watermark
  • Credits Never Expire
  • Failed Generations Auto-Refunded
RECOMMENDED

Standard

30 Credits
$9.99$14.97Save 33%

Best value for creators producing high-definition dance videos.

Up to 2 Videos15 Seconds · 720P HD
  • 30 Credits
  • Up to 2 Videos
  • 15 Seconds · 720P HD
  • 5s & 15s Duration Supported
  • AI Video from 1 Photo
  • Reference-guided Choreography
  • Fast Processing Queue
  • No Watermark
  • Credits Never Expire
  • Failed Generations Auto-Refunded

Party Pack

90 Credits
$24.99$44.91Save 44%

Great pack for testing multiple outfits, poses and dance choreography.

Up to 6 Videos3 HD Videos · 15s
  • 90 Credits
  • Up to 6 Videos
  • 3 HD Videos (720P · 15s)
  • 5s & 15s Duration Supported
  • Full HD 1080P Available
  • AI Video from 1 Photo
  • Reference-guided Choreography
  • Fast Processing Queue
  • No Watermark
  • Credits Never Expire
  • Failed Generations Auto-Refunded

Crew Pack

300 Credits
$69.99$149.70Save 53%

Maximum credit volume for creators making multiple dance videos.

Up to 20 Videos10 HD Videos · 15s
  • 300 Credits
  • Up to 20 Videos
  • 10 HD Videos (720P · 15s)
  • 5s & 15s Duration Supported
  • Full HD 1080P Available
  • AI Video from 1 Photo
  • Reference-guided Choreography
  • Maximum Value for Multiple Videos
  • No Watermark
  • Credits Never Expire
  • Failed Generations Auto-Refunded

Pricing & Video Capacity Notes:

  • Comparison strikethrough prices are calculated based on the Starter single video ($4.99 / 5s) baseline.
  • Maximum video counts are estimated based on standard 15-second videos at 480P SD quality (15 credits/video). Higher quality settings (720P HD at 30 credits, 1080P Full HD at 60 credits) use more credits, reducing the total video count accordingly. The Starter pack includes 1 video at 5s (480P only).