How to make a Bulin 47 AI video from one photo

The whole process takes a few minutes and needs exactly one photo. Here is every step in detail, what to expect from the render, honest comparisons with other ways to do it, and how to fix the usual problems.

Updated October 12, 2026

The three steps

Make a free account

Sign up with Google or a 6-digit code sent to your email - no password to invent. The account keeps your clips in your library and lets you buy a pack when you want one. It takes under a minute - Google is the fastest path, the emailed code works everywhere.

Upload one photo

Pick a clear, front-facing photo of one person - JPG, PNG or WebP up to 10 MB. The model works from what it can see, so the better the photo, the better the clip. You will also confirm you have the right to use it before the render starts.

Get your clip

The render usually takes two to four minutes and runs on our side, so closing the tab does not lose it. The clip lands in your library automatically - download the vertical MP4 and share it straight from the file - it is a standard MP4, so it plays everywhere.

Picking the right photo

What to expect

Clips are 8 or 15 seconds in 9:16, 1:1 or 16:9, from 480p up to 1080p, with the original freestyle sound and no watermark. Plan on a few minutes of waiting, not hours. A failed render is refunded to your clip balance automatically - you never pay for a clip that did not finish.

Three ways to do it

There is more than one way to get a Bulin 47 clip. The honest trade-offs, before you spend time or money:

Troubleshooting

• Face not matched? Use a brighter, sharper, more frontal photo - sunglasses, hats and heavy side angles are the usual culprits.
• Hands look odd? AI still struggles with fingers sometimes. Re-render with a different photo, or post the clip anyway - most viewers never notice on a phone screen.
• Upload rejected? Check the file is a JPG, PNG or WebP under 10 MB, at least 300 px on each side, and that it actually opens in an image viewer.

Quick answers