You, on the verse
Stand on the right of the hanging mic, nod through the hook, then take the verse. The camera cuts to your face for the featured close-up.
58th Floor AI Video Generator
Two people. One hanging mic. Take the verse, add a friend for the hook, or let AI create your partner in the teal studio.
Take your spot at the mic
15 SEC · 720POne photo works: an AI-created partner takes the hook. JPG, PNG or WebP · up to 10 MB each.
Checking your account…
One request makes one 15-second, 9:16 MP4. One or two photos use the same 1 credit. Confirmed failures return the credit.
By creating, you agree to our Terms and Privacy Policy.
CHOOSE WHO TAKES EACH PART
Your photos decide the lineup
The 58th Floor AI effect is built around a handover, not two people performing the same line. The first singer finishes the hook. Then the other turns toward the silver mic, takes the verse and gets the close-up.
You do not need to film yourself, record vocals or write a prompt. Add a clear photo for each person you want in the studio. Keep the photos in the right order: changing their order changes who gets the featured verse.
Your friend, on the hook
Take the left spot and finish the opening hook. When the verse starts, listen, smile and move to the beat. Leave this photo empty for a fictional AI-created partner.
INSIDE THE 15-SECOND CLIP
Four shots. One handover.
The reference video supplies the studio and edit. These are the intended moments; precise gestures and timing can change during generation.
- 0–5s
The hook, together
A two-shot frames both performers. Your friend finishes the hook while you nod and clap, then turn toward the mic.
- 5–8s
Your verse close-up
The camera cuts in as you lean toward the mic. Photo 1 remains the featured performer through the cut.
- 8–11s
The studio goes wide
Both people come into view in the teal room. You perform; your partner supports the moment.
- 11–15s
Back to the two-shot
The edit returns to both performers, with natural hand gestures and your partner moving to the beat.
ONE PHOTO STILL MAKES A DUET
No friend photo? Take the verse anyway.
The one-photo version keeps the same two-person structure. You supply the verse performer, and the model creates a fictional adult for the other spot. It is a different partner, rather than a second copy of you.
Use two photos when you want a particular friend, sibling or partner on the hook. Both people must agree to appear. Each photo should contain one adult; group pictures make the role assignment less clear.
Both versions use one video credit. Choose your lineup ↑
FROM PHOTOS TO YOUR MP4
How to make a 58th Floor AI video
Add your performer photos
Use a front-facing, evenly lit photo for the verse. Add the hook performer if you want a specific friend. Waist-up framing helps preserve the outfit and gives the model room for the gestures.
Choose the soundtrack
Reference sound adds the template recording after the visuals are generated. Silent delivers an MP4 with no audio, so you can choose a track in your editor or social app.
Create with one credit
Sign in, confirm photo permission and click Create. The task opens in My videos, where you can follow progress, review your inputs and download the finished clip.
A CLEARER START
Photos that work with the close-up
Keep the face readable
Photo 1 receives the closest shot. Pick a sharp face with even light, visible eyes and no heavy filter. Avoid dark images, sunglasses, hands over the face and screenshots with existing captions.
Give each person their own photo
Different hair and outfits help separate the two roles. Keep one adult in each image and avoid overlapping faces. Each file can be JPG, PNG or WebP, between 1 KB and 10 MB.
Leave room for the performance
A waist-up image is a stronger starting point than a very tight selfie. AI still creates the unseen body and studio, so clothing, proportions and hand details can change. Review the result before sharing.
BEFORE YOUR SESSION
58th Floor AI questions
One shared credit, one video. See the credit packs or contact support.
What is the 58th Floor AI video?
It is an AI rap duet made from photos, recreating a two-person mic handover in a teal studio. Your partner finishes the hook, you lean toward the hanging microphone for the verse, and the edit moves between your close-up, a wide shot and the two-shot. The scene is generated from your photos and a video reference.
How do I use the 58th Floor AI filter?
Upload a verse performer photo, optionally add a friend for the hook, choose reference sound or silent, and create with one video credit. The effect is rendered as a downloadable MP4 after submission. You can post that file in your preferred social app.
Can I make it from one photo?
Yes. Photo 1 is enough. You take the verse and the close-up, while the model creates a different fictional adult for the hook. A second photo gives you control over that partner's appearance. The cost is one video credit either way.
Who gets the close-up?
Photo 1 always takes the verse on the right side of the mic and gets the close-up. Photo 2 takes the hook on the left. To give your friend the verse, put their photo first and yours second before creating.
Can I choose a silent video?
Yes. Choose Silent before submitting. Reference sound adds the recording accompanying the template; Silent removes the audio completely, ready for a track you add in your editor or social app. The soundtrack is an existing recording, not a clone of your own voice.
How long is it, and what does it cost?
One request creates one 15-second, 720p, vertical 9:16 MP4 for one shared video credit. Use the same credit balance as Bulin 47 AI and the other tools. Buying a credit pack is separate from starting a video, and adding the optional friend photo does not cost another credit.
Will the faces and lip movements be exact?
Your photos guide the appearance, and the reference guides the shots and timing. The model creates new frames, so faces, hair, outfits, hands and mouth movements can vary. Check both performers through the cuts, especially the verse close-up. Exact likeness and word-for-word lip sync are not guaranteed.
Is this a free 58th Floor AI generator?
Watching the template previews is free. Creating your own video requires Google sign-in and one paid video credit. There is no free render or daily generation allowance for this tool.
What if generation fails or I close the page?
You can leave after submission and follow the existing task in My videos. A confirmed generation failure returns the video credit. If submission cannot be confirmed, check that task and contact support before trying again. Queue times vary, so the task page is the place to check progress.
Your close-up is the next part.
Make your 58th Floor video ↑GenBulin47AI is an independent service. The video is an AI-generated remake, not an official performance or artist endorsement.