Google has launched Nano Banana 2.1, a new image generation and editing model based on its Gemini 3.6 Flash technology, with improvements in image quality, instruction following and editing capabilities. The company says the model also significantly reduces generation costs, with API image prices now roughly half those of its predecessor.
Nano Banana 2.1 can generate images from text prompts and edit existing visuals using natural-language instructions. Google says users can modify elements such as clothing, backgrounds, objects and visual styles without recreating an image from scratch. The model can also combine multiple reference images while maintaining consistency across the generated result.
The new model supports up to 14 reference images and can maintain consistency for as many as four characters and 10 objects. This could help users creating a series of visuals involving the same people, products or objects. Google has also improved its handling of text within generated images, making the model better suited to posters, diagrams, infographics and other visuals that require readable text.
Also Read: Woman Missing For Two Years: Allahabad High Court Orders CBI To Trace Her Whereabouts
Nano Banana 2.1 supports 1K, 2K and 4K image outputs and offers different levels of AI reasoning depending on the amount of processing required. Another addition is Search grounding, which allows the model to use information from Google Search and Google Image Search when generating images based on real-world information.
The biggest change is the pricing for developers using the API. Google says a 1K image generated with Nano Banana 2.1 costs about $0.0336, compared with $0.067 previously. A 4K image costs around $0.0756, down from $0.151, bringing the cost of image generation to roughly half the previous level.
Nano Banana 2.1 is being rolled out across several Google products, including the Gemini app, Google AI Studio, Gemini API, Google Search’s AI Mode, Google Ads, Flow and Stitch. The lower pricing could be particularly significant for developers and businesses generating large volumes of images, while the improved editing and reference-image capabilities expand the model’s use for both individual creators and commercial workflows.
Also Read: Property Dispute Claims Life In Maharashtra; Man Shot Dead, Five Including Grandfather Arrested