The fastest way is right here: one photo in, the full 19-second freestyle out. Below is how the trend works, the other tools people use, and how to get a result that actually looks like you.
Front-facing, sharp, one person, with hair, clothes and upper body visible. Selfies work; tight face crops and sunglasses don’t.
Keep the original street crowd or turn it into a World Cup crowd. 480p costs 10 credits, 720p HD 22.
We replace the performer in Kevin Wai’s clip with you across all 19 seconds and keep the original sound. Download the MP4, or the silent copy if you prefer to add the sound inside TikTok.
| Method | What you need | Trade-off |
|---|---|---|
| CapCut template | A video of yourself to drop in | Edits your own footage; it does not put your face into the original scene. |
| Dreamina / Seedance prompt | The original clip, a clear photo and a written prompt | Flexible, but you prepare the clip and tune the prompt yourself. |
| Higgsfield Genjutsu | The original clip and character images | Popular for celebrity remixes; credit-based and queues can be long. |
| Bulin 47 AI (this site) | One photo | Clip, prompt and sound are pre-set; you only choose crowd and quality. |
Tool details change often; this table reflects what tutorials showed in early October 2026.
The download keeps the original freestyle audio and is 9:16, so it uploads to TikTok, Reels and Shorts without cropping.
If a platform mutes the audio for copyright reasons, post the silent copy and pick the trending Bulin 47 sound inside the app — it lines up from the first second.
CapCut templates are free but only edit your own footage. Recasting yourself into the original scene with AI costs credits here: one 480p video is 10 credits ($9.99 on its own).
19 seconds — the part of Kevin Wai’s clip used by the most-liked AI versions, starting right as the camera turns to the performer.
No. The clip, timing and sound are already set up — you only upload your photo.