Google has made its image model, Gemini 2.5 Flash Image, also known as Nano Banana, generally available for production use. According to the company, new features include support for ten different aspect ratios for various formats, from cinematic to vertical, and the option to specify image-only output.
In a post, Google provided several tips for users to get the most out of the model, which is integrated into the Gemini app.
- Maintain consistency: The model can reuse characters and scenes across multiple images. This allows for altering outfits, poses, or backgrounds while preserving the subject’s likeness.
- Edit details: Users can change specific elements within an image using natural language. For example, a prompt like “the dog’s mouth is closed” can alter a photo without affecting the rest of the scene. This also works for changing the color of an object.
- Explore creatively: Instead of complex prompts, the model can interpret simple, conversational instructions. It also allows for combining up to three separate images to create a new composition.
For developers, the model is accessible through the Gemini API on Google AI Studio and Vertex AI. Google highlighted apps created with the tool, such as a virtual fitting room and an animator that turns images into GIFs, to showcase its capabilities.