
Hook (left)
Stands on the left, a step from the mic. Finishes the hook with the raised-hand gestures.
Replaces the left performer in the opening bars and the hook.
58th Floor AI · the teal studio trend
Two photos, 15 seconds, one hanging microphone
Upload one or two photos and get a 15-second vertical clip of the teal-studio performance: one silver microphone hanging between you, the left seat finishes the hook, the right seat takes the verse, the song plays underneath. No prompts, no editing.
A single continuous shot — no cuts to stitch. The camera holds a medium frame and drifts in while the two of you trade lines. Both seats are optional: one photo gets you the hook and leaves the original performer in the verse.

Hook (left)
Face to camera, head and shoulders, even light.

Verse (right)
Face to camera, head and shoulders, even light.
The teal wall, the hanging mic and the same slow push-in every time. Only the people in the frame change.



58th Floor AI is the photo-to-video version of one specific scene: Cordae and Anderson .Paak performing “58th Floor” on A COLORS SHOW — a solid teal backdrop, a single silver condenser microphone hanging from the ceiling, two performers a step apart trading bars.
The original is one continuous take, which is what makes it a good template. There is no cut to fake, no location to rebuild and nothing to lip-sync by hand: the camera holds a medium frame and pushes in slowly while the two of them trade the hook and the verse.
Our studio keeps the set, the hanging mic and the song, and replaces the two performers with the people in your photos. One photo takes the hook and leaves the original version of the verse in place. Two photos cast both seats. You get 15 seconds, vertical, with the original audio underneath, in a few minutes.
Two uploads and one button. There is no prompt box, because the scene is already decided.
The first photo takes the left seat and the hook. The second takes the right seat and the verse. Leave the second empty and the original performer keeps it.
480p, 720p or 1080p. The credit cost sits on the button before you pay, and it does not change with the number of photos.
The clip renders as one take, gets trimmed to exactly 15 seconds, and arrives as a vertical 9:16 MP4 with the song already in it, no watermark.
Left goes first. That is the whole casting decision — the seat you pick decides which bars your face carries.

Hook (left)
Stands on the left, a step from the mic. Finishes the hook with the raised-hand gestures.
Replaces the left performer in the opening bars and the hook.

Verse (right)
Stands on the right and takes the verse, turning half a step toward the mic.
Replaces the right performer once the hook lands.
The clip is a medium shot of two people from the chest up, so half-body portraits work better than tight headshots.
Same portrait rules as every other effect on the site. Tiny Desk AI photo requirements
15 seconds is the longest single take we render, and it is the most reference-heavy: 14 credits at 480p, 24 at 720p, 38 at 1080p. One or two photos cost the same.
| Cast | 480p | 720p Best value | 1080p |
|---|---|---|---|
| 1–2 peopleSame price for one photo or two | 14credits | 24credits | 38credits |
Priced per finished video, never per seat. 1080p is an AI upscale of the 720p master.
Credit packs: 7 credits for $6.99, 16 for $14.99, 50 for $44.99. Credits last 12 months. Pricing
The rest of the site explains the format, the photo rules and the other effects.
Pick the hook seat first — that is the face the clip opens on.
Make my 58th Floor video