Turn Two Photos into the Smalltown Boy AI Scream-and-Swap Video

Cast two photos into the viral Bronski Beat Smalltown Boy routine: one screams into the fisheye lens, the other dances behind, they swap, then break into the 80s duo dance. Pictell edits the real 20-second trend clip frame by frame, so every beat lands where the song drops.

Seedance 2.5 Edit bills input plus output seconds at the selected resolution. The estimate uses the 21s classic length, and the 20s vertical cut costs slightly less.

Result

Your generated result will appear here.

Cast the Duo, Keep Every Beat

Most Smalltown Boy AI tools animate two still photos and hope the dance looks close. Pictell starts from the actual trend video and only swaps who is in it, so the scream, the swap, and the duo dance stay on cue.

Pick Who Screams First

Photo 1 opens the clip pressed against the lens, hands on head, mid-scream. Photo 2 dances in the background, then swaps to the front. Choosing the order is half the joke.

Choose the Cut and the Scene

Pick the vertical 9:16 cut for TikTok, Reels, and Shorts or the classic 3:4 backyard original. By default the duo leaves the small town for Times Square, and you can rewrite that line to send them anywhere.

Remix the Pair in Canvas

Open the full workflow to recast either role, try a cat and its owner, compare both template cuts side by side, and keep the version your group chat replays most.

A Real Video Edit, Not a Guessing Animation

A Real Video Edit, Not a Guessing Animation

The trend only works when the timing is exact: the fisheye scream, the hand slapped toward the lens, the swap, and the duo dance as the run away, turn away chorus kicks in. Pictell runs Seedance 2.5 Edit on the original clip, so movements, gestures, timing, and camera motion are kept while only the two people are replaced.

The Small-Town Boy Finally Makes It to the Big City

The Small-Town Boy Finally Makes It to the Big City

Smalltown Boy is a song about leaving home. That is why our default scene moves your duo from the backyard fence to a daylight Times Square, which makes the meme land twice. Prefer your own hometown, an office floor, or a neon 80s club? Change one line of the prompt.

Any Two: Besties, Couples, Cats, Anime Duos

Any Two: Besties, Couples, Cats, Anime Duos

The funniest versions pair two subjects who should never do this together: you and your cat, two coworkers, a couple, a founder and their mascot, or two anime characters. Each role gets its own upload, so neither face gets blended into the other.

Two Template Cuts, One Workflow

Two Template Cuts, One Workflow

Post the 20-second vertical cut natively on TikTok, Reels, and Shorts, or use the 21-second 3:4 classic that the trend started from. Both keep the original Smalltown Boy audio, and Canvas lets you render both from the same pair of photos.

Smalltown Boy AI FAQ

What is the Smalltown Boy AI trend?
It is a two-person scream-and-dance meme set to Bronski Beat's Smalltown Boy. One person screams into a wide fisheye lens with their hands on their head, the second dances behind them, they swap places, and the clip ends with both doing a retro 80s dance. People use AI to put friends, couples, pets, or anime characters into the routine.
Is Smalltown Boy a new song?
No. Smalltown Boy is the 1984 debut single by the British synth-pop band Bronski Beat, sung in Jimmy Somerville's falsetto, and it reached No. 3 in the UK. The song is about a young gay man leaving a hometown that rejected him. Short-video templates brought it back in 2026, and this AI trend is fan-made, not an official Bronski Beat campaign.
How is Pictell different from other Smalltown Boy AI generators?
Many generators animate two still photos into a new dance, so the moves and timing drift from the meme. Pictell edits the real trend clip with Seedance 2.5 Edit and only replaces the two people, which keeps the scream, the swap, the camera moves, and the original audio on cue. You also choose which photo plays which role, pick a vertical or classic cut, change the background with one prompt line, and open the whole workflow in Canvas.
Which photo should go in Photo 1 and which in Photo 2?
Photo 1 is the first screamer, filmed close to the lens, so a clear face matters most. Photo 2 dances in the background first and needs a visible full body. For the biggest laugh, put the calmer or more serious subject in Photo 1.
Can I use cats, dogs, or anime characters?
Yes. Any clear subject with a visible face works, including pets, mascots, and anime or cartoon characters. Mixed pairs like you and your cat are some of the most shared versions of this trend.
Why does the default scene move to Times Square, and can I change it?
It is a nod to the song: the small-town boy leaves home for the big city. The last line of the prompt sets the background, so open Advanced options and replace Times Square with any place you like, or ask the model to keep the original backyard.
Can I change the background?
Yes. Edit the background part of the prompt to drop the duo somewhere else and make the clip more fun. You can also open the workflow in Canvas and switch to another model if you want to lower the cost.
Should I pick the vertical or the classic cut?
Choose Vertical 9:16 (20 seconds) for TikTok, Instagram Reels, and YouTube Shorts. Choose Classic 3:4 (21 seconds) if you want the original backyard-fence framing that made the trend recognizable.
Does the video keep the Smalltown Boy music?
Yes. The workflow keeps the audio of the template clip instead of generating new sound, so the scream and the swap stay synced with the song.
How many credits does it cost?
Seedance 2.5 Edit bills the input and output seconds at the selected resolution. The button shows the estimate before you generate. 480p is the cheapest option and is plenty for a meme post.
Is this affiliated with Bronski Beat?
No. This is an independent Pictell workflow for making AI-generated trend remixes. Only upload photos you have permission to use, and label realistic outputs as AI-generated when you publish them.

Run away, turn away, and post the remix

Upload two photos, choose who screams first, and get the full scream-and-swap routine on beat. Then keep recasting the duo in Canvas.