SuperAPI AI Directory

Pixo

Pixo (Bulin 47 AI) is a browser-based AI face-swap video tool that replaces a single photo uploaded by the user into the viral freestyle videos of Bulin 47 and Kevin Wai. After the user uploads a clear frontal photo, the AI model follows the original footage frame by frame, preserving the rap movements, handheld camera work, crowd, and original dembow sound effects. It supports 5, 10, 15 seconds and the full 19 seconds, 9:16 vertical format, 480p to 1080p, and can be used in the browser on iPhone, Android, or a laptop without installing an app. It uses one-time credit packs, starting at $4.99, and credits are refunded for failed videos.

Open website

About this tool

Pixo (the page name is Bulin 47 AI) is an AI video remake tool that runs in the browser, whose core purpose is to "swap" a photo uploaded by the user into the viral street freestyle video of Dominican rappers Bulin 47 and Kevin Wai. The original was filmed outside a restaurant in Madrid, with Kevin Wai holding a round portable speaker playing a dembow beat, and Bulin 47 wearing sunglasses and a pink mesh shirt freestyling. This tool lets users, friends, or even pets replace the rapper in the original footage while preserving the original movements, laughter, the moment of raising hands, handheld camera work, crowd, and original sound effects.

In terms of main capabilities, the tool accepts a single photo of one person that is clear, frontal, and with an unobstructed face, and the AI video model remakes it frame by frame following the original footage. The user's photo contributes the face, hair, skin tone, and the clothing and accessories visible in the photo; pets or cartoon characters retain their own appearance and complete the movements. Kevin Wai, the round speaker, the red light entrance, the lighting, and the handheld camera in the original remain unchanged, while the background crowd is redrawn as fictional people. Video duration can be selected as 5, 10, 15 seconds, or the full 19 seconds: 5, 10, and 15 seconds are the opening portion of the freestyle and can be generated by all models; the full 19 seconds is only available on Seedance 2.5. In terms of resolution, Seedance 2.0 and Seedance 2.5 support 1080p. Model choices include Seedance 2.0 Mini, Seedance 2.0 Fast, Seedance 2.0, and Seedance 2.5: Seedance 2.5 is the most stable for faces and hands during fast motion, Seedance 2.0 and Fast are in the middle, and Mini is the low-cost preview option. The output is a 9:16 vertical MP4 with the original sound, ready to post directly to TikTok, Reels, or Shorts without cropping or re-exporting.

There are three steps to use it: first, upload a single photo of one person that is clear, frontal, and with an unobstructed face; showing hair and upper body helps with clothing transfer, and pets and cartoon characters can also be used; second, choose the duration and model; third, wait a few minutes, and the video will appear on the page and in the history, where you can save the MP4 to your album and post it. The tool runs in Safari, Chrome, or Samsung Browser on iPhone, Android, and laptops, with no app installation required.

In terms of pricing and credits, all packages are one-time payments, credits never expire, and failed or rejected videos are not charged. A 15-second 480p Seedance 2.0 Mini video consumes 8 credits; the $4.99 Starter package can pay for one, and larger packages can reduce the per-video cost to about $0.53. Seedance 2.0 Fast, Seedance 2.0, and Seedance 2.5 are more expensive, while 5-second or 10-second videos are cheaper. A full 19-second Seedance 2.5 consumes 84 credits at 480p and 188 credits at 720p. The packages listed in the documentation include: Starter (10 credits, one-time, 1 15-second 480p video, watermark removed, refund on failure); the 150-credit package (26 + 124 bonus credits, 18 15-second 480p videos, supports 5 to 19 seconds, four Seedance models, up to 1080p); the 400-credit package (60 + 340 bonus credits, 50 15-second 480p videos, priority support); the 1,200-credit package (158 + 1,042 bonus credits, 150 15-second 480p videos, aimed at meme pages and agencies). Any package removes watermarks from all videos.

The main use case is personal social trend videos: users turn themselves, friends, or pets into Bulin 47 freestyle versions and post them to TikTok, Instagram Reels, or YouTube Shorts labeled as AI content. The original audio belongs to its owner, and platforms may mute it or claim copyright, so it is recommended to add the trending Bulin 47 sound from within TikTok or Instagram.

Limitations and notes: this tool is an independent fan tool, unrelated to Bulin 47, Kevin Wai, and their team, and is not produced, endorsed, or sponsored by them. The template only recreates one person (the rapper), and Kevin Wai's position remains as is; couples or groups need to crop photos separately and create them individually, as group photos usually cause face blending. Commercial use is not recommended because it involves a real artist and their music; commercial use requires permission from the rights holders, and many advertising platforms reject AI portrait content, so videos generated here are for personal non-commercial posting only. Do not upload photos of children, celebrities, or public figures, and do not upload photos of people without their consent.Photos should have even lighting, include the head and upper body, and show the eyes; avoid sunglasses, masks, face coverings, group photos, heavy filters, or small distant faces. Lip movements follow the original rap, close but not fully frame-accurate, and the fastest hand movements may briefly blur. Photos are stored in private storage bound to the account and are only sent to the model that creates the video; they are not publicly displayed or used as examples. Deletion can be requested by emailing support@bulinaivideo.org from the account email. The page mentions that "there are currently no free credits," and the Starter package ($4.99) can pay for one 15-second 480p video; it also mentions that free credits can only pay for Seedance 2.0 Mini videos, and that before the first purchase videos carry a small moving watermark. These two pieces of information are inconsistent.