Uptown Funk AI Prompt: Five-Person Dance Guide
Last updated:
A useful dance prompt separates who each person is from what the reference video supplies. Use the example below as a starting point for a five-person crew, then check your platform’s reference syntax and supported settings.
Copy and adapt the English prompt
Use the reference video for the choreography, camera framing and timing. Place Image 1 as the center lead, Image 2 at far left, Image 3 at far right, Image 4 at inner left and Image 5 at inner right, as viewed by the audience. Keep each person’s facial identity and position consistent across cuts. Follow the group’s street dance; keep feet grounded, hands natural and bodies distinct. Avoid duplicated faces, merged limbs and subtitles.
This is an editable example based on the role and reference instructions in AI Uptown Funk's code, not a guaranteed recipe or a transcript of every provider request. Attach one portrait per person and a suitable reference video where your chosen tool supports them. Reference labels and controls vary by platform.
Map the five identities before generating
Photo 1 is the center lead; photos 2 and 3 are the outer left and right; photos 4 and 5 are the inner left and right, from the viewer’s perspective. For fewer people, fill the slots in this order. The number selected is the number of identities replaced, not a promise that the reference contains no other dancers.
Use one clear, evenly lit adult portrait per person, with their permission. Distinct, unobstructed faces make it easier to assess the result. Keep the photo order fixed when comparing attempts.
Use Kling 4.0 · 720p on AI Uptown Funk
Select Kling 4.0 · 720p in the generator: the current site option requests 16:9, 15 seconds and 720p for 149 credits. Select the cast size and outfit, then upload one photo per slot. The site assembles its own provider instructions; there is no custom prompt field in this generator.
The reference video supplies the movement and framing. The site restores the reference audio after generation. These are settings for this site, not a claim that every Kling interface accepts the same references, duration or parameters.
Diagnose a result before changing the prompt
If two people exchange faces after a cut, review the full video and check the original slot order. Try clearer, distinct portraits and keep the same cast order. If hands merge or feet slide, compare the problematic shots against the reference; a request for natural hands and grounded feet is guidance, not a guarantee.
For a controlled comparison, change one variable at a time: portrait, outfit or model. Keep the others fixed and review wide shots, close-ups and transitions. Each completed attempt costs credits; a failed task returns the credits charged for that task. These are troubleshooting suggestions, not documented results from a paid test.
Create with the built-in reference
To use AI Uptown Funk, open the generator below, choose a model and outfit, and upload the portraits in order. The site handles the reference instructions for you. Download and inspect the finished clip within the retention period, and disclose that it is AI-generated when sharing.
Only use photos and reference material you have permission to use. Check the rights to the audio and video before publishing or using the result commercially.
Your crew. The Uptown Funk routine.
Create my crew video