
Good starting point
- One adult per photo, with a clearly visible face
- Front or slight side view, with the whole head in frame
- Even light and natural, recognizable facial details
- Head and shoulders visible, with a natural expression
Three photos. Suspect #1, Suspect #2 and the detective.Your cast, the familiar singing scene. Add three photos for Suspect #1, Suspect #2 and the detective.
The Brooklyn Nine-Nine singing lineup, with a different cast.
Five numbered suspects take turns singing “I Want It That Way,” with cuts to the observation room and reactions. AI remakes follow that familiar sequence with a different cast. This template assigns your three photos to Suspect #1, Suspect #2 and the speaking detective. The reference clip guides their performance and the edits.
NBC / Brooklyn Nine-Nine: the original singing sceneThis is a fixed three-role scene remake, not a song generator or full-cast editor. The other suspects and the woman in the observation room are not replacement targets.
The recognizable scene is the starting point; the new casting is the twist.
Brooklyn Nine-Nine turns a police lineup into a call-and-response performance, interrupted by reactions from the observation room.
NBC / Brooklyn Nine-Nine: the original singing sceneTechnology-personality and football remakes use matching shots and singing turns, showing how the scene can be recast.
Downey: technology-personality remakeAssign a portrait to Suspect #1, Suspect #2 and Detective, then choose the length and resolution. The remaining cast is outside the replacement scope.
Three photos, three fixed roles and your choice of video settings.
Add one clear adult portrait for each role: Suspect #1, Suspect #2 and Detective. Match each photo to its role thumbnail.
Pick 5 or 15 seconds, then 480P or 720P. Review the total credits on the generation button.
Start the generation, follow its status and preview the result. Check the full clip before downloading or sharing.
Use one clear portrait for each role, so the model can distinguish the three people across the shots.


The identity changes; the scene should still feel familiar.
Keep each photo assigned to its suspect number or the detective, even when the camera cuts.
Check that the three faces stay recognizable without swapping identities during singing, reactions and side views.
Keep the numbered cards, singing turns and observation-room reactions in the right order.
Review mouths, hands, audio and nearby faces before sharing. One good frame is not enough.
The generation button shows the total credits for your selected settings before you submit.
Select 5 or 15 seconds at 480P or 720P. The default is 15 seconds at 480P. Longer clips and higher resolution use more credits.
Each new submission is a separate generation. Check all three photo assignments and the completed clip before starting another attempt.
AI may alter facial details, clothing, color, numbered cards, hands or mouth movement. The example is one result, not a guarantee for every photo.
These are different workflows. A CapCut template does not automatically support the same three role assignments.
Try another fixed performance or bring your own reference clip.
Upload three clear adult portraits, one each for Suspect #1, Suspect #2 and Detective. Choose 5 or 15 seconds and a resolution, then review the credits and generate. Preview and download the result from the same tool.
The three fixed roles are Suspect #1, Suspect #2 and the male detective speaking in the observation room. Each gets a separate photo. The other three suspects and the woman in the observation room are configured to remain unchanged; this is not full-cast replacement.
The three portraits guide the assigned identities in a fixed reference scene. Outfits, skin tone and fine details can also change, so review the full result. This is not a manual outfit editor or a guaranteed full-body copy.
Choose 5 or 15 seconds at 480P or 720P. The short cut gives less screen time to each role; it is not the opening of the 15-second cut. Preview the selected reference to see who appears and for how long. The default is 15 seconds at 480P, and the generation button shows the total credits for all three roles before you submit.
The template uses the selected reference clip's audio rather than asking you to create a new song or voice. Review the result for audio timing and mouth movement. Generating a video does not grant rights to the underlying song, dialogue or footage.
Crop each image to one adult with a visible face before uploading. Check the numbered thumbnails for Suspect #1 and Suspect #2, and the observation-room thumbnail for Detective. If the assignment is wrong, replace the photo in that slot before generating; do not rely on left-to-right positions across different shots.
A task marked failed follows the site's existing credit-return process. Check the task status before submitting another generation. A completed result with imperfect likeness or lip-sync is different from a technical failure.
No. This is an independent template, not affiliated with or endorsed by Brooklyn Nine-Nine, NBC or the song’s rights holders.