Google launched Nano Banana 2.1, a new image model that generates images for nearly half the cost. Based on the Gemini 3.6 Flash model, it follows instructions more accurately and handles edits better. Users can now create visuals using up to 14 reference images while keeping important elements consistent throughout.
Google has launched Nano Banana 2.1, its latest AI model for generating and editing images. The company says the new model can create better-looking visuals, follow instructions more accurately and handle image edits better than its predecessor. And then there is the cost. According to Google, Nano Banana 2.1 can generate images for nearly half the price of the previous model.
Nano Banana 2.1 is based on Google’s Gemini 3.6 Flash model and is being rolled out across Google’s products, including the Gemini app, Google AI Studio, Gemini API, Google Search’s AI Mode, Google Ads, Flow and Stitch.
So, what’s new with Nano Banana 2.1?
Like previous versions, users can simply describe the image they want in a text prompt and the model will generate it. But Google says the new version is better at following detailed instructions and keeping important elements consistent when an image is edited.
For example, users can give the model several reference images and ask it to combine them into a new visual. Nano Banana 2.1 can work with up to 14 reference images at once and maintain consistency for up to four characters and 10 objects. This could be useful when creating a series of images where the same person, product or object needs to appear across multiple visuals.
The model can also edit existing images using text instructions. Users can ask it to change a person’s clothes, alter the background, add or remove objects, or change the overall style without having to recreate the entire image from scratch.
Google says Nano Banana 2.1 also improves the way AI-generated images handle text. This means it should be better at creating posters, diagrams, infographics and other visuals where readable text is important. It supports 1K, 2K and 4K image output, while users can also choose different levels of AI reasoning depending on how much processing they want the model to use.
Another notable addition is Google’s Search grounding. Nano Banana 2.1 can use information from Google Search and Google Image Search when generating an image. This could be useful when users want to create visuals based on real-world information.
Nano Banana 2.1 price per image
The biggest change, however, could be the price. A 1K image generated through the API now costs around $0.0336, down from $0.067 previously. A 4K image costs about $0.0756, compared with $0.151 earlier. In other words, developers can generate images for roughly half the previous cost.
For developers and businesses generating large numbers of images, that price difference could add up quickly. Nano Banana 2.1 is therefore not just an update to Google’s image-generation model; it also makes generating images at scale considerably cheaper.
The model is now rolling out across Google’s products, giving both developers and regular users access to its new image-generation and editing features.