Bulin 47 AI with Kling, CapCut, ChatGPT or Gemini: Step by Step

Which tool does what for the Bulin 47 trend

All of these can get you close to the trend. They differ in how much you have to write, how many attempts you pay for and whether the clip has a voice:

• Kling: photo plus prompt, newest models generate sound with the picture. Most control, most retries.
• CapCut and Dreamina: Dreamina (CapCut’s AI site) makes the clip with Seedance; CapCut is where you edit, add text and pick a sound.
• Gemini: photo-to-video with sound in the Gemini app, up to 10 seconds, on a paid Google AI plan.
• ChatGPT: writes the prompt, the verse and the caption, but does not make the video.
• Higgsfield: many viral versions came from here; see how the Higgsfield Bulin 47 videos are made.
• This site: one photo in, a finished 15-second freestyle with sound out, no prompt.

How to make the Bulin 47 AI video in Kling

Kling is the tool people name most often under these clips. Kling 4.0 is its headline model, sold on audio-visual performance, and Kling 3.0 already generates lip-synced speech with the picture. The route:

• Open image-to-video and upload one sharp, front-facing photo of yourself or of an adult who agreed. Never a photo of Bulin 47, Kevin Wai or any celebrity.
• Paste a scene prompt: a night street outside a red-lit restaurant, a friend holding a portable speaker, a crowd cheering, the person rapping in Spanish over a dembow beat, phone-camera look. Our Bulin 47 AI prompt template is free to copy.
• Set 9:16 vertical, turn native audio on and choose the longest length your plan allows.
• Write the verse yourself, in your own words. Do not paste the original lyrics: they belong to the artist and many apps block them.
• Render, watch with sound and retry. Expect a few attempts: lip sync, hands and the crowd are where clips usually break.

How to make it with CapCut and Dreamina

CapCut on its own does not put you into a new scene. Its templates add text, cuts and effects to footage you already have. The AI part lives in Dreamina, CapCut’s AI site, which runs ByteDance’s Seedance video models:

• In Dreamina, choose image-to-video with Seedance and upload your photo.
• Use the same scene prompt as above. Audio comes with the video; lip-synced speech is a paid-plan feature.
• Download the clip and open it in CapCut.
• Add the caption on screen. Most people use "dame hombro", the line everyone knows. On TikTok, pick a sound from the app’s own library.
• Export in 9:16 and post.

Can ChatGPT make the Bulin 47 AI video?

No. ChatGPT does not render video. OpenAI closed the Sora app on April 26, 2026 and the Sora API on September 24, 2026, so there is no OpenAI route to a finished clip any more. Answers that send you to Sora are out of date.

ChatGPT is still useful for the parts around the video:

• Turn our prompt template into a version with your own twist (a different outfit, a beach instead of a street).
• Write an original four-line Spanish verse about you, to paste into Kling or Dreamina.
• Write captions and hashtags, or translate "dame hombro" for friends.
• Check a finished clip against the list below before you post.

Can Gemini make it?

Partly. The Gemini app now uses Gemini Omni, which replaced Veo 3.1 there. It turns up to five photos into a video with generated sound, but clips stop at 10 seconds and it needs a paid Google AI plan. Ten seconds fits the opening of the freestyle, not the whole 15-second trend clip. Videos carry Google’s invisible SynthID watermark.

Check your clip before you post, whichever tool you used

• The mouth moves with the words, not a beat behind.
• Hands stay hands: five fingers, no melting during gestures.
• The crowd reacts at the right moments and does not morph into each other.
• The face still looks like the person in the photo.
• Nothing in it comes from the original clip: no footage, no original audio, no real artist’s face.

When the one-photo generator is simpler

If you want the trend clip and not a prompting project, the generator on this page has the scene, the moves and an original beat built in. Upload one photo, pick 6 or 15 seconds, and you know the price before you start: 100 credits with Standard 480p or 220 credits with HD 720p (credit packs start at $3.99 for 40 credits). Failed renders are refunded. For a side-by-side comparison of every route, including face swaps and filming it yourself, see Bulin 47 AI alternatives.