The car driving AI video trend is the newest wave to break on Reels and TikTok: AI-generated cinematic in-car shots of celebrities — or of you — behind the wheel of a luxury car. The signature look is a Mercedes on a night drive, neon city lights sliding across the windshield, rain beading on the glass, shallow depth of field doing the rest. Six creators riding this format have already pulled a combined audience in the million-view range, and the prompt recipe is simple enough that a first clip takes minutes, not hours.
New here? The homepage overview covers the last wave in one screen, and the AI video trends hub tracks every meme as it breaks.
The format is a short, film-like clip of a person driving an expensive car — most often a Mercedes at night — that no camera ever filmed. Generators produce the whole scene from one photo plus one paragraph of prompt text, so the driver can be a famous face, a fictional character, or your own selfie. Prompt-sharing posts circulate constantly on TikTok and Instagram, which is why the same neon-and-rain look keeps appearing in different accounts' feeds.
It is the same anatomy as the Hotel Lobby AI wave: one recognizable visual, one tool moment, and a participation hook where the output stars you. The difference is the entry cost — no reference video needed, just a portrait.
Kling AI is the tool most tutorials and creators use for this meme. The flow is image-to-video: you upload a photo, describe the driving scene, and the model animates it into a five-to-ten-second cinematic clip. A full walkthrough of the workflow is in this YouTube tutorial.
A second common path runs through ImagineArt with the Seedance model. Creators who use it report the same recipe — photo in, cinematic drive out — so treat it as a substitute if Kling's queue or credits are tight, not as a different trend.
Both tools can also generate the scene with no photo at all: describe the driver and the car in text and let the model invent the person. You lose the "that's actually me" hook, which is the engine of the meme, but it is the fastest way to test a prompt.
This is the community's popular template shape — swap the variables in brackets to make the clip yours:
cinematic shot of a person driving a Mercedes at night,
neon city lights, rain on windshield, shallow depth of fieldChange one variable at a time so you can see what each edit does:
For more ready-made driving prompts, Media.io maintains a dedicated car video prompts library that follows the same structure.
Yes. Kling AI's free credits cover your first short generations, and each clip in this trend is only five to ten seconds, so a free allowance goes a long way before any paid plan matters.
Five to ten seconds. That is the length every major example uses — long enough to sell the cinematic mood, short enough to loop and re-watch, which is what feeds the algorithm.
Reels and TikTok, with trending audio attached. The trend spread through those two platforms and the sound choice is a large part of reach; posting without audio leaves most of the wave on the table.
No, but it helps. The selfie route is what makes viewers stop scrolling — "wait, that's him in a Mercedes" — while pure text-to-video produces a stranger in the driver seat.