Google’s new AI image generator took my shirt off

I gave Google New Nano Banana Pro Try, and immediately took off my clothes. I didn’t ask for it, but apparently the AI ​​model decided that my greeting card would look better with more skin.

Nano Banana Pro, as the name suggests, is aimed at professionals. Powered by Gemini 3it’s actually an upgrade of the company’s popular photo creation and editing tool that has gone viral on social media Turn selfies into highly realistic 3D sculptures. Google says it lets you create higher-quality images that you can print, provide legible text on images, and blend multiple images together into a single composition. This app is also for “people who want to feel like professionals,” said Naina Raisinghani, product manager at Google DeepMind. Edge. That sounds good, because I’m by no means a professional. For me, the results were shiny, but goofy. It looked good, but felt amateurish.

Using the Nano Banana Pro is very simple: you go into the Gemini app, select “Create Photos,” and then switch to “Thinking” mode. Simply plug in the prompt (and photo, if you’re using one) and get started. It’s also free, though there are limits, with quotas expanding for Google AI Plus, Pro, and Ultra subscribers.

Google makes some bold claims, promising “studio-quality designs,” “flawless text rendering,” and a host of stylish and creative tweaks. To test this, I uploaded a simple photo of myself nearby Edge Office in New York with the Brooklyn Bridge in the background. I asked Gemini to change the lighting from day to night and she did a very good job. The result seems believable. It even handled details that often bog down image generators, such as getting cars to go in the right direction. Adjusting the camera angle was equally easy. I asked Jiminy to recreate the shot as if it were taken from a higher right angle, which she did.

Google also says the Nano Banana Pro can create charts and graphs to help visualize real-time information like weather or sports. Being British, I asked about the weather for the next four days in Washington, DC, and New York City, where I am currently located. Visually, the infographic would have been right at home on a basic forecasting site. The text and numbers looked natural—a far cry from the distorted nonsense you often see in AI-generated images—and Gemini gave me a list of citations at the end that helped me confirm their accuracy.

The model stumbled a bit on more complex tasks. I asked him to summarize the latter edge A story about how Europe is reducing AI and privacy laws In comic book form. The images and text were indeed presented in flawless cartoonish font, but the storyboard didn’t summarize the story at all, giving a vague overview of the story. Block Artificial Intelligence Law instead of. Maybe the problem was because I gave Jiminy a link to the story, instead of pasting the text into it.

It gave me a passable synopsis in comedic style when I did so. It conveyed the gist of the actual story, although I don’t think I would have been able to understand it as easily had I not written the source material. It also consists of phrases that did not appear anywhere in my article.

To really feel like a professional designer, I tried my hand at making greeting cards. Christmas is coming, after all. Considering that I only uploaded three selfies, Gemini honestly did a great job creating three full-body versions of myself, each in different outfits and with different facial expressions. I also created a realistic snowy environment with Christmas trees, as I asked her to, and attached the words “Merry Christmas!” Above as you requested.

Gemini took liberties by asking them to change the snowy background of the card to a summer beach for an Australian-style holiday. Those freedoms were my fake clothes: two of my clones were naked. It was strange. There were also some notable ones Feet created by artificial intelligence And a smiling sandman to replace the snowman from the winter scene (which was built by topless lookalikes). However, there were problems – the sandman lacked shade, unlike the other creatures shown in the photo, and the Christmas lights in the palm trees glowed magically in the bright sun. I tested his meticulous editing skills by asking him to add some power to just one copy, which he did in seconds (if only it were that easy in the real world). Overall, the quality was great, and the photo was fairly believable (except for the abs) if you didn’t know there was a large tattoo missing on my chest.

It wasn’t all great, though. The form failed to maintain the exact text on my card that I requested of it. Instead of “Happy Birthday!” She chose “Australian Summer Christmas!” She also seems to have a conflict with animals: my sister’s cat is sitting in exactly the same pose as the reference photo I’ve provided in each copy of the card (she’s been given a whimsical Santa hat, though).

Overall, I was impressed. The Nano Banana Pro is a clear upgrade over the base model. I’ve been able to dial in more precise edits, and it actually produces legible text, removing a huge barrier that prevents generative AI tools like this from being usable in the real world. But, unfortunately, these features were not enough to make me a good designer.