Google makes Nano Banana generally available
Google Cloud has made Nano Banana 2 and Nano Banana Pro generally available. Technically named Gemini 3.1 Flash Image and Gemini 3 Pro Image, the models are intended for image generation and editing in business workflows.
The models are available via the Gemini Enterprise Agent Platform. Google emphasizes enterprise infrastructure, security, and SLAs in this regard. Developers can also use the models via the Gemini API, but according to Google, that route is not covered by the same enterprise SLA.
A new feature is that Nano Banana 2 can also use videos as input in preview. The model can then analyze visual context, subjects, and actions from video footage to create new images, such as thumbnails, infographics, or contextual visuals.
The 1K and 2K output capabilities are generally available for both models. 4K output remains in preview for the time being.
Google positions the models primarily for creative production at scale. Adobe uses Nano Banana models in Firefly and GenStudio. WPP integrates them into WPP Open for agentic marketing workflows. Shopify mentions applications for product photography and lifestyle imagery, while URBN deploys them for early product development. Magnopus uses Google’s generative media in professional production environments with 3D workflows.
This is important for companies because image generation is shifting from standalone creative experiments to built-in production workflows. Brands want to create variants faster, personalize campaigns, and adjust product images without losing control over brand consistency.
The move also shows how Google is increasingly linking multimodal AI to enterprise platforms. Images, video, text, and agents are moving closer together in the same cloud environment. This makes generative media more practical for companies but also increases the importance of rights management, brand control, and clear review processes.