What if you could edit photos just by saying what you want? This is how “Nano Banana” works in Gemini, the new Google jewel
Imagine uploading a photo of yourself to an app and simply saying, “put me on the beach in a cowboy hat”… and boom! The image appears exactly as you imagined it. You didn’t use Photoshop, you didn’t crop anything, you didn’t even move a finger beyond typing that sentence. This is no longer science fiction: it’s Nano Banana, the new AI editing tool that Google has integrated into its Gemini app. And yes, it’s getting a lot of buzz.
What is Nano Banana and why all the fuss?
Developed by the Google DeepMind team, Nano Banana is a generative artificial intelligence model that takes image editing to the next level. What is revolutionary is not only the quality of the edits, but the ease with which you can achieve professional results using natural language.
Want a practical example? You can upload a photo of your cat and ask: “Put some dragon wings on it and place it in a medieval tower at sunset”. The model not only understands what you say, but maintains the identity of the subject (yes, your cat will still look like your cat) and delivers a coherent and realistic image.
Why does this tool make a difference?
Here are a few reasons why Nano Banana is surprising the technology community:
- Precise editing with natural language: You don’t need to know about editing or learn weird commands. Just type what you want as if you were talking to a friend.
- Consistency in faces and objects: Has it ever happened to you that in some AI editors, your face changes slightly from one image to another? Here, that doesn’t happen. The model keeps your features even if you change background, outfit, or artistic style.
- Step-by-step editing (multi-turn): You can start with a base image and make successive changes: “paint the wall red”, “now add a sofa”, “put a lamp in the corner”. Like having a creative assistant at your disposal.
- Mix of styles: Do you like a flower print? You can use it to design a dress or decorate a piece of furniture. The model understands styles and applies them naturally.
- Watermarking and SynthID: All generated images are clearly identified as AI-created, thanks to visible and invisible tagging technologies, bringing transparency in a world increasingly full of synthetic images. Remember, I have a post discussing SyntID on my blog.
Let’s see a quick example in which I upload to Google AI Studio, where I use the functionality to create a first image as a model and then iterate it in several situations.

Now, using a simple prompt, I ask him to position me on the beach wearing a Google T-shirt (as indicated in a second iteration with the model) and also holding a giant banana in his hands. As you can see, he keeps faithfully to the character but makes the changes I have asked him to make without a fuss.

Finally, I ask for something completely different to see how the model behaves, and with a simple prompt, I tell the model to place me in front of the Burj al Kalifa in the city of Dubai dressed in an Emirati outfit, ale-hop…. magic 🙂

How does it compare to other tools?
Nano Banana does not arrive in a vacuum. Competition is fierce, with OpenAI’s DALL-E integrated into ChatGPT, Midjourney, Stable Diffusion, and the advances of Adobe Firefly. But where these tools still have limitations, especially in maintaining a person’s identity across multiple images or enabling iterative editing, Google has the edge.
In addition, everything is integrated within Gemini, Google’s multimodal app, which now allows not only chatting with AI, but also creating, editing, animating, and more images. All in the same place.

This feature can be used by everyone! From free users to those with a subscription. All you need is a photo and a creative idea.
In my case, I have a paid subscription, and I must comment that from the usual Gemini interface, it has not allowed me to upload an image (indicating functionality not yet available). However, I have managed to do so from the link I provided above from Google AI Studio.
Google has not only launched an eye-catching feature, but has also democratized high-quality visual creation. Tools like Nano Banana turn anyone into a creative editor, no technical knowledge required. From designers prototyping ideas, to users who simply want to see their dog dressed as a superhero, creativity no longer has technical limits.
What crazy image would you like to generate with Gemini?
Have a good week!
RELEVANT SOURCES:
Nano Banana! Image editing in Gemini just got a major upgrade | Google Enhances Gemini AI Image Capabilities to Challenge ChatGPT