Hop Hop Hop AI: How to Make the Squash-and-Swap Trend From Photos
Make the TikTok Hop Hop Hop gag with AI: the six takes, one or two photos, who hops first, the 8 seconds beat by beat, and what it costs.
The joke is a camera trick. Someone holds a flat palm in the air right in front of the lens, and a friend standing far behind looks tiny underneath it. Every time the palm comes down, the friend sinks lower, three hops on the "hop hop hop" of the sound, then a hard cut swaps them for the payback. The Hop Hop Hop template recreates that with your photos.
1Upload one photo per friend
Photo 1 starts in front, photo 2 hops far behind; after the cut you swap. With one photo, the friend comes with the template.
JPG, PNG, WebP or HEIC · under 20 MB0/2 photos
Add 1 to 2 photos, one each, in the order of the slots.
Photos that work best
- Clear, waist up

- Screenshots, stickers

- Head only

- Too dark or blurry

2Pick a template
- 8 s video
- 9:16, like the original
- Original song available
- No watermark
3Version
By making a video, you confirm you own these photos or have permission from everyone in them, and they don't show celebrities or anyone under 18.
Renders in about 5–10 minutes. We email you the link. Automatic refund if the photo is rejected or the render fails.
Where the trend came from
The sound was posted on TikTok on 24 September 2026. On 18 October a best-friends duo posted the squash-and-swap version that passed 3.7 million views, and within a week the "squished or get squashed" take and an AI version with two fairy-tale warriors had millions more.
The six takes
On the template page you can pick which viral clip to follow: Squished (a hallway), Best friends (a bedroom), the emo duo in a living room, Night square, a pet take where your dog or cat squashes you back with a paw, and AI heroes with a goblin and a giant warrior. All six take one or two photos.
What happens in the 8 seconds
- 0–2.5 s: photo 1 is right at the lens; photo 2 is far behind.
- 2.5–4.3 s: the palm goes up over photo 2, who hops three times and sinks into a crouch.
- 4.3 s: hard cut. Photo 2 is at the lens now and photo 1 is the one far behind.
- 4.3–8 s: the palm comes down again and photo 1 gets squashed back.
Tips for the photos
- Full-length or waist-up. The person behind is seen head to toe while hopping.
- Clear faces, one person per photo, no sunglasses.
- Swap to switch roles. Whoever is photo 2 gets squashed first.
- Ask your friend first. Use only photos of people who are in on the joke.
What does it cost?
Balanced is 10 credits at 480p and 20 at 720p; Premium is 20, 40 and 70 credits at 480p, 720p and 1080p. Without an account you can buy a single video at checkout. No watermark, and failed renders are refunded automatically.