Back to blog

AI Rap Video Prompts: Photo and Scene Preparation Guide

RapVideoGo

Prepare reference photos and write a clear AI rap video prompt. Learn image order, cast labels, scene directions and practical fixes for crowded short clips.

AI Rap Video Prompts: Photo and Scene Preparation Guide

A useful AI rap video prompt makes the cast, lyrics and scene easy to follow. Reference photos describe appearance; the prompt explains who performs, what they say and how the shot should work. Adding more directions is only helpful when those directions agree.

Use this guide before opening the rap video studio, or when revising a brief after a result feels crowded.

Choose reference photos for a specific job

Use a sharp photo with even light, a visible face and minimal obstruction. For two performers, separate single-person images are easier to label. Large sunglasses, a hand across the face or a face at the edge of a group can make the intended identity less clear.

The original studio supports up to nine JPG, PNG or WebP reference images, with an 8 MB limit per image and a 24 MB total. Those limits are capacity, not a recommendation to fill every slot. Start with the fewest references that explain your cast.

The battle studio uses exactly two performer slots and asks for one clear adult face in each photo. Choose images of yourself or people who agreed to appear.

Match reference order to your role labels

@1 means the first uploaded image, and @2 means the second. Define the labels once:

A is the person in @1, standing on the left. B is the person in @2, standing on the right. Exactly two performers; keep their identities distinct.

Then use A and B consistently in the lyrics and actions. If you remove or reorder references, reread the labels before generating. A group reference can contain several people in the original studio, so explicitly identify the intended cast instead of assuming every reference means one singer.

Use five parts in the prompt

  1. Cast: who appears and which reference identifies them.
  2. Scene: one location and a few important objects.
  3. Lyrics: short lines with speaker labels when needed.
  4. Sound: language, delivery and broad beat direction.
  5. Camera and actions: a simple shot and manageable gestures.

For example, here is a starting scene direction for an office duet:

Two coworkers in a bright office after a long meeting. A is @1 on the left and B is @2 on the right. One microphone between them; a coffee cup on the desk. A steady front-facing two-shot with both faces visible. Relaxed hip-hop beat and clear English vocals. A delivers the first line, B answers, then repeat A and B. The listener nods rather than singing over the lead. Perform only the supplied lyrics.

Add the verse below that paragraph. Try the rap video ideas article if you still need a situation to write about.

Remove directions that compete

“A single unbroken two-shot” conflicts with “cut between four close-ups.” “Exactly two performers” becomes unclear if a later sentence introduces a cheering crowd. Resolve the conflict in the text before generating.

For a short clip, prioritize one setting, one exchange and one ending. Long verses compete with pauses, reactions and elaborate gestures for the same few seconds. If the last line keeps getting rushed, cut words before adding more timing instructions.

What to change after reviewing a result

  • Faces merge or swap: simplify the references and restate the cast.
  • Both people appear to lead together: assign the lines and describe the listener's reaction.
  • Words are skipped: shorten the verse or choose a longer available duration.
  • The camera loses a performer: ask for a steady two-shot with both faces visible.
  • The scene feels busy: remove extra locations, props and movements.

These are ways to make the brief clearer, not guaranteed fixes. AI generation can still vary. Change one major part at a time and keep a copy of the earlier prompt so you can compare results.

Final check before generating

Confirm photo order, speaker labels, language, duration, resolution, frame shape and the displayed credit cost. The studio generates AI vocals and a beat rather than using an uploaded recording. Play the finished result before downloading or sharing.

Continue with the full rap video workflow, the two-person battle tutorial or the birthday lyrics examples.