Google announces Nano Banana 2.1 image model

Google announced Nano Banana 2.1 on October 6, presenting the image-generation and editing model as an update with improved visual design, more precise mask-based edits and stronger consistency across successive changes. Reports say it is rolling out across the Gemini app, Google Search’s AI Mode, Google Ads and developer tools, while Google Cloud documentation lists it as generally available globally on the Gemini Enterprise Agent Platform. The model is designed to generate images from text and edit supplied images. Google says it can accept up to 14 reference images at once and maintain consistency for as many as four characters and ten objects. Outputs are available at 1K, 2K and 4K resolutions. Developers can also adjust a reasoning setting and use Google Search or Google Image Search to ground image generation in real-world information, according to the reported product details. Google’s developer API rates were reported at $0.0336 for a standard 1K image, compared with $0.067 for Nano Banana 2. A 4K image was listed at $0.0756, down from $0.151. Reports also described a further 50% discount for batch processing.