Immy Dorse 
It’s not every day you see a new internet trend take over so completely. But here we are again. Google’s latest AI experiment, a tool they’re calling Nano Banana, has millions of people turning their selfies into miniature, collectible-style figurines. Honestly, the results are kind of spooky. They look like something you’d pull right off a shelf in a specialty toy store.
How It All Started
The feature officially launched in late August as part of Gemini. Google’s team posted a simple tutorial on X, and it just exploded from there. I think it hit a nerve. Within a matter of weeks, users had generated over 200 million images. The Philippines went absolutely wild for it, creating over 25 million on its own. A Google VP even mentioned they’d gained 10 million new users specifically for this one trick. That’s a lot of digital dolls.
It’s actually the third big AI portrait trend this year. Back in March, everyone was using ChatGPT to make themselves look like characters from a Studio Ghibli film. Then in April, the “Barbie Box” challenge had LinkedIn professionals turning themselves into action figures. This one feels different, though. The detail is just incredible. The AI nails facial features, clothing textures, even dynamic poses. People keep getting fooled into thinking they’re looking at a real product photo.
Not the Only Game in Town
Of course, Google isn’t alone now. Alibaba’s new model does something very similar, and some say it handles tricky poses even better. For the techies who want total control, there are open-source options like Flux Kontext. That one doesn’t have any content filters, which is a double-edged sword, really. It’s powerful, but the responsibility is all on you.
Making Your Own Figurine
So how do you do it? It’s surprisingly simple and free. You just go to the Gemini website or open the app. Look for the banana emoji—you can’t miss it. Upload a clear, well-lit photo of yourself or a friend. Then, you paste in a specific prompt.
The one that seems to work best is about creating a “1/7 scale commercialized figurine” on a desk with a clear base and a computer screen showing 3D software next to it. It’s very specific, but that’s what makes it work.
A few tips: ask for “PVC” or “resin” to make it more realistic. Use a photo where you’re in an action pose, not just standing still. And if the face doesn’t look quite right, you can add “the face remains unaltered” to the prompt. Sometimes that helps. Sometimes it makes it a bit too real and loses the doll effect.
You get a limited number of tries for free each day. Other platforms offer unlimited generations, but the quality might dip a bit after a while.
More Than Just a Gimmick
This isn’t just for goofing around anymore. Politicians in Asia are using these AI figurines in their campaigns. On TikTok, people are building entire dioramas with multiple figures, sharing them under #NanoBanana. It’s become a whole creative subculture almost overnight. It makes you wonder what’s next. Or how long until we’re all just collections of digital miniatures.