AI video tools are popping up everywhere these days, and most of them promise the same thing. That means turning a simple image into something more dynamic. I’ve seen people use them for everything from anime-style clips to bizarrely realistic conversational avatars.
But one trend I kept coming across was a little more unexpected. It’s an AI kissing video.
So I decided to try it myself.
This idea seemed simple enough
This concept is wild in a very Internet sense.
You take two regular photos (usually portraits), upload them to an AI tool, describe the scene, and it generates a short video of two people interacting. In this case…well, it’s a kissing scene.
There is no editing software. No animation skills. Just a photo and a prompt.
It felt like it could either work amazingly well or be completely cursed.
First try: Not great, not terrible.
I chose two very ordinary photos.
- Front facing
- decent lighting
- nothing too dramatic
Upload them, type something like “Soft Romantic Moments” and press generate.
result?
To be honest…it’s somewhere in between.
It wasn’t perfect, but it wasn’t a complete mess either. The movement seemed a little stiff in places, but I could clearly see what it was trying to do. It definitely felt more like a “moment” than just two still images.
It’s also a little creepy. But in some ways, AI is promising.
Second attempt: much better
Made some small adjustments.
- I chose a photo with a more similar angle.
- I used a simpler prompt
- I avoided anything too dramatic.
It actually made a huge difference.
The second version seemed much smoother. The transition between the two faces felt more natural, and the entire clip was easier to watch. It’s obviously not Hollywood-level realism yet, but it’s definitely something I can see people sharing as short-form content.
Why is this suddenly popping up everywhere?
After trying it out, I can kind of understand why this format is trending.
It accomplishes several things at once.
- It’s emotional (even if a little awkward)
- Create quickly
- Works perfectly for short-form feeds
- Feels “weird enough” to get attention
You don’t need a complete video idea. All you need is two images and a concept.
That’s a pretty low barrier.
the tools behind
There are a lot of tools currently experimenting with this kind of thing. Some are general AI video generators, while others focus on specific effects or scenes.
What struck me was how simplified the process was. Tools like this AI Kissing Video Generator take away most of the hassle. Basically, you can go from idea to output in minutes.
That’s a big part of why these clips are going viral so quickly.
It’s not perfect (and that’s part of its charm)
The truth is, it’s still AI.
sometimes:
- Her facial expression seems a little off.
- Movement is not completely smooth
- Results vary depending on the photo
But to be honest, that unpredictability is part of what makes it fun. Some feel amazingly sophisticated, while others feel a little unrealistic, and both types tend to get noticed online.
final thoughts
I went into this expecting something gimmicky, but it’s actually a very good example of where AI content is heading.
It’s fast. It is accessible. And turn something as simple as a photo into something people actually want to see.
Platforms like Lanta AI are clearly leaning in that direction, emphasizing rapid, idea-driven content creation rather than complex editing.
Can I use it forever? Probably not.
But as a simple and creative experiment?
Well…this is definitely one of the most interesting things I’ve tried recently.
