Bulin 47 AI Freestyle Video Generator

Where the Bulin 47 freestyle came from

In early October 2026, Dominican dembow artist Bulin 47, born Hancel Teodoro Vargas Reynoso, was heading into a restaurant in Madrid when Colombian creator Kevin Wai started a beat on a small portable speaker. Bulin turned around and freestyled on the spot, bouncing his shoulders hard on the beat and pointing at the speaker while people crowded around and filmed. Kevin Wai posted it, and by 7 October media reports put the views in the tens of millions; Spanish sports paper MARCA covered it on 8 October. The moment turned into an official song, "Dame Hombro", released by Bulin 47 and Kevin Wai with producer Breyco En Producidera.

The AI version keeps the setting everyone recognizes and swaps in new people: you as the freestyler, and a friend as the one hyping you up with the speaker. That is what this page does with your photos.

What happens in your clip

The preset builds a fresh scene rather than editing the original video. It is night on a city street, outside a restaurant entrance glowing with red neon. A tight crowd presses in, many of them holding up phones. Right beside you, your friend from photo 2 holds one round black portable speaker toward you, bouncing and cheering; with only one photo, the friend is just a hand in the foreground. The camera is handheld and a little shaky, framed from the waist up, and it follows you as you bounce your shoulders, point, lean toward the lens and deliver the hook. Both of you keep your own faces, hair and clothes.

The default video is 12 seconds, 720p, vertical 9:16, on Seedance 2.0 Mini, and costs 10 credits. Our smallest pack is US$9.90 for 10 credits, so one pack makes one video, and you can pay as a guest. The sound is generated with the picture: a loud dembow beat, a Spanish-style rap vocal and crowd cheers. It is not the original freestyle or "Dame Hombro", and it is not your voice. There is no watermark, and failed or timed-out renders are refunded automatically.

Tips from our own test renders

• One speaker, no logo. In an early test the model put two speakers in the shot, one with a brand logo. The preset now asks for exactly one round black speaker with no logo, and the renders we kept follow it. One render still turned it into a tall cylinder speaker, so small changes like that can happen.
• Neon letters are made up. The red signs above the door often show partial, invented lettering. That is the model filling in a restaurant sign, not a real business name.
• A sharp face matters most. You lean toward the lens and the camera stays close, so blur or heavy shadow on your face shows up clearly.
• Any outfit works. Our test renders used a grey hoodie and a floral summer dress, and both came through as they were. You do not need the pink mesh shirt from the original.
• Two people, two clear roles. In our two-photo test the friend stood at the rapper's side holding the speaker, the freestyler stayed in the middle of the frame, and the two faces did not mix. Put the person who should do the rapping in slot 1.
• The opening frame can be your photo. In one of our test renders the very first moment is the uploaded photo before the cut to the street, so avoid a photo you would not want shown.

Who to cast as the freestyler and the friend

The joke works best with someone who would never freestyle in public: a calm parent, the quiet coworker, a friend on their birthday. Put them in slot 1 and yourself in slot 2 as the one holding the speaker, and it becomes a clip about you two. Ask them first, and post it with a caption that sets up the contrast.