Here’s how people make 3D figurines of these viruses

Each major new AI image generator appears to be causing a new virus trend that will take over the internet for several days. The GPT ‑ 4o was released in March, so it was a Studio Ghibli style image. Google’s Nano Banana, also known as Gemini 2.5 Flash Image, is a 3D figurine.

It’s not a completely new trend. People were already showing a tendency to use GPT-4O to create their own action figures. However, the nano bananas appear to be able to provide more detailed, refined and quick compliance. Sometimes they’re cute and sometimes a little creepy. People have also discovered that static images of 3D figurines can be converted into videos.

What are the trends in AI 3D figurines?

Google’s release of Gemini 2.5 Flash Image, known as Nano Banana, has sparked the latest trend in AI 3D figurines. Google itself has driven trends by posting images generated by models on social media.

People can upload photos of themselves or become another Gemini target and provide text prompts with instructions to generate images of 3D figurines. With the right prompt, AI models can generate images that combine the portraits of reference photos with detailed toy figure appearances such as props, accessories, packaging, and more.

Most of the time, people are generating images in which 3D figurines are placed on a desk in front of a computer screen, displaying the toy design in one of the best 3D modeling programs.

How do people make AI 3D figurines?

Although various AI models can be used to try to generate images that look like 3D figurines, people are mostly using Google’s Gemini 2.5 flash images (Nano Banana).

Click “Try Nano Banana.” You will be given a box to enter the prompt. Here are the prompts Google suggests:

Create 1/7 scale commercialized figurines of photographic characters in a realistic environment, in a realistic style. The figurine is placed at the computer desk. The figurine has a round, transparent acrylic base, with no text on the base. The content on the computer screen is the 3D modeling process for this figurine. Next to the computer screen is a toy packaging box designed in a style reminiscent of high-quality collectible figures with original artwork printed on it. The package features two-dimensional flat illustrations.

This kind of thing makes you realize how few photos of yourself, especially full-length photos. The only thing I found was a relatively low resolution photo taken under extremely harsh conditions. So, although the nanobananas didn’t have great data to work on, the model still identifies the subject of the photograph with a peer and tripod and transformed it into a 3D figure.

You might like it

Having a better quality image can give you more detail and a more realistic portrait. You can also add instructions to the prompt to specify whether you want the figurine to be in a different pose. You can also specify the text to display in the package. It also does a pretty decent job by applying it instead of the regular AI garbled text. Like all AI image generators, results may be inconsistent.

You can animate images generated by AI using any number of AI video generators that appear.

Do you want to be more creative and do it in the old fashioned way? Don’t miss this year’s illusion of competition entry of the year, which actually recreates AI figurine trends.

For more information about AI News, don’t miss Openai’s big bet story on AI filmmaking using AI animation Critterz.