7 surprising uses of Nano Banana 2, the ultra-fast image generating AI from Gemini

The generation of images through AI has become one of Gemini’s flagship features, given its appeal and ease of use for both professional users and amateurs alike. In order to further drive this burgeoning success, Google has just introduced Nano Banana 2, its image generation model that offers high fidelity and speed.

Nano Banana 2 is based on the Gemini 3.1 Flash Image model, and combines the best of both worlds by delivering the intelligence and quality of Nano Banana Pro with the swift performance of the Flash models. This innovation is now available in the Gemini application, Google Search, AI Studio, Google Cloud, Google Ads, and Flow.

Using Nano Banana 2 in Gemini is extremely straightforward, as this model has replaced Nano Banana Pro by default in the “Fast,” “Reasoning,” and “Pro” modes, meaning that you will not need to make any manual adjustments beyond selecting the “Create Image” tool. Now, without further delay, let us explore the capabilities of this AI!

1. Create Images with Preset Styles

The first major innovation you will encounter when selecting Gemini’s “Create Image” tool is the style menu. Now, the AI provides a gallery of preset styles that you may apply to your creations to ensure they align even more closely with the concept you have envisioned.

Of course, if you wish, you may opt not to select any style and instead provide your own instructions.

2. Create Infographics by Leveraging Web Information

Nano Banana 2 is capable of extracting information from Gemini’s real-world knowledge base and can also access Google Search in real time to retrieve relevant information to enhance your designs. In this way, creating infographics becomes more straightforward while also improving the accuracy of the results. For example, I requested that it create an infographic about the life cycle of the dandelion flower.

It is important to note that Nano Banana 2 provides the sources from which it has extracted the necessary data to generate the infographic, thereby enabling users to consult them. Access to web information reduces the likelihood of technical errors or inaccurate information being represented; however, this is not an absolute guarantee, and we recommend thoroughly reviewing the information presented.

This capability can also prove highly useful for transforming notes into diagrams or generating visual representations of data.

3. Maintain consistency across up to 5 characters and 14 objects

Google’s AI for image generation has enhanced its ability to preserve the appearance of key elements when generating images based on protagonists. When it comes to subjects or characters, Nano Banana 2 can maintain the consistency of up to five characters within a single workflow. Meanwhile, for objects, the achievement is even greater: it is able to preserve the fidelity of up to fourteen objects.

This capability allows you to generate original narratives, whether creating a story that you have invented or producing sketches for storyboards.

4. Change the background while retaining quality

Nano Banana 2 enables you to change the background of an image while keeping key elements or central characters intact and seamlessly integrating them with the new setting.

5. Generate solid variations by modifying context, not essence

A major breakthrough offered by Nano Banana 2 is that it enables you not only to change the setting of an image, but also its context and even relevant visual information pertaining to its primary subject. For instance, you can select an image of a small sign containing drawings of local birds and flowers and generate a variation styled for India, displaying native wildlife and flora, with the text in Hindi. The composition and essence are maintained, but you receive a faithful and professional variation.

6. Merge images to create new narratives

The capabilities for consistency and contextual variation also enable you to experiment with Nano Banana 2, creating something entirely new by merging several images. In my case, after generating several images featuring friendly muppets traveling through Italy, I created an image of a photo album and then attached all of these images in Gemini to request that they be combined to form a travel album chronicling the muppets’ adventures.

And this was the result:

7. Use multiple aspect ratios and resolutions

With Nano Banana Pro, you were already able to generate images in different aspect ratios, but now you will enjoy greater control over these dimensions, as the new model responds with greater precision to your requests and manages image dimensions more effectively.

Likewise, you may also specify the final resolution of your creation, from 512 pixels up to 4K, ensuring that your work appears sharp whether for a vertical Instagram Story or a panoramic desktop wallpaper.

Verification with SynthID watermark

As with its predecessor, Nano Banana 2 integrates Google’s SynthID technology. This means that every image generated by this AI is marked with a digital watermark that is invisible to the human eye, yet detectable by Google’s verification systems.

However, even though humans are unable to see this watermark, we are able to identify it by using the Gemini app. Thus, if you suspect an image has been generated by Nano Banana 2, you may upload it to the Gemini chat and inquire about its origin. The AI will examine the file, detect the SynthID watermark, and promptly inform you of its findings.

Photo: Image generated with Nano Banana 2