What happens in the bob filter video, second by second?
Haircut video creates a new 10-second phone video from your photo, set in your photo's location, showing your face and outfit. The sequence follows the template clips that define the trend:
| Time | What happens |
|---|---|
| 0-2 s | You glance at the front camera with an uncertain frown and raise scissors beside your long hair. |
| 2-4 s | You gather a long section to one side and snip straight across at jaw level. The cut hair falls away. |
| 4-6 s | You trim the opposite side. A blunt, chin-length bob appears, and you gasp in surprise, hand covering your mouth. |
| 6-7.5 s | A brief spray of water over the fresh bob. |
| 7.5-10 s | You sweep your hands through the bob, toss it and smile at the camera. |
How Smew AI makes it: your photo feeds directly into Seedance 2.5, ByteDance's video model, with no intermediate image, preserving facial accuracy. Hair color and part stay the same; only length is altered. The output is vertical, 720p, with its own audio: scissor sounds, a gasp and laughter, no background music.
Bob photo feeds your photo to Nano Banana 2.1, Google's image model, modifying only the hairstyle. You receive one 2K vertical image of yourself sporting a fresh bob.
The create page displays the exact cost on the Generate button before you commit, and each generation produces a unique result.
