Two-photo trend

Hop Hop Hop AI

Squished or get squashed

Two photos, one hard cut. One of you points and presses a flat palm down; the other hops, folds and sinks. Then you swap. Eight seconds, 9:16, no editor, no prompt.

  • 2 photos
  • 8 seconds
  • 9:16 vertical
  • 720p or 1080p
  • Trend sound included
  • No watermark

1 Add both photos

0 / 2

One photo per person. The trend only works as a duo — one presses, one gets squished, then you swap.

The pointer

Stays original

The bouncer

Stays original

Both photos are required. Add one for each person.

2 Pick the quality

720p is the native render. 1080p is an AI upscale of that same master.

This video: 10 credits · ready in a few minutes

Made from these two photos · 8s · 720p

The exact clip you get

This is a real render from the studio below, made from the two photos next to it. Nobody filmed anything.

Photo 1
Photo 1
Photo 2
Photo 2

Both seats are required. A one-person version breaks the joke, because there is nobody to squish and nobody to swap with.

How the Hop Hop Hop AI video is made

No timeline, no keyframes, no editing app. The action is fixed; you supply the two faces.

  1. 01

    Add two photos

    One per person. The first photo takes the close-up; the second starts down the hallway. Both people appear in both halves, so both photos matter.

  2. 02

    Pick 720p or 1080p

    720p is the native render and the fastest. 1080p is an AI upscale of that same master, for feeds that re-encode hard.

  3. 03

    Download and post

    An 8-second vertical MP4 with the trend sound already laid under it, no watermark. Sound and picture are locked to the same beats, so posting it does not drift.

The two roles in the duo

The format is a two-person camera trick. Whoever stands at the lens holds a flat palm over the person far down the hallway and presses down; they crouch lower and lower. The palm never touches anyone — they are metres apart and only line up from the camera. At the hard cut they trade places and it happens the other way round.

The pointer

The pointer

Starts at the lens. Points back down the hallway, raises a flat palm over their head and presses them down.

Takes the close-up in the first half; after the hard cut they are the one down the hallway.

The bouncer

The bouncer

Starts far down the hallway. Hops on the beat, then folds and sinks as the palm comes down.

Starts down the hallway in the first half; after the hard cut they take the close-up.

Both seats are required. A one-person version breaks the joke, because there is nobody to squish and nobody to swap with.

What photos work for Hop Hop Hop AI

The model has to keep both faces straight across a hard cut, so the photos do most of the work.

Works

  • One adult per photo, face to camera, unobstructed.
  • Full body or knees-up framing — the sink needs visible legs on the far person.
  • Even light, no heavy filters, no sunglasses.
  • Both photos of the same kind: if one is a headshot and the other a full body, the swap reads unevenly.

Breaks it

  • Group photos, or two people in one frame.
  • Screenshots, stickers, or faces covered by a hand.
  • Minors, celebrities, or anyone who has not agreed to appear.

Upload only photos of adults who have agreed to appear. Label the result as AI-generated when you post it.

Hop Hop Hop AI pricing

One price per finished video. Credits never expire, and a failed render returns them automatically.

720p

$9.99

10 credits

Native render · fastest

1080p

$16.99

17 credits

AI upscale of the 720p master

Credits come in packs; the per-video figure above is the single-pack price. Bigger packs bring it down. Pricing

Hop Hop Hop AI questions

Keep reading

The trend explained, the two roles in detail, and what to fix when a render comes out wrong.

Hop Hop Hop AI

Two photos, one hard cut. One of you points and presses a flat palm down; the other hops, folds and sinks. Then you swap. Eight seconds, 9:16, no editor, no prompt.

Add your two photos