What is Nano Banana? Here's why Google's image editing AI model is going viral

Google has expanded its suite of artificial intelligence tools with the launch of Nano Banana, an image editing model integrated into the Gemini app.

The model, officially called Gemini 2.5 Flash Image, is designed to enhance visual editing through natural language prompts, enabling users to make advanced modifications quickly and precisely.

Advanced image editing features

Nano Banana brings several new capabilities to AI image editing. A standout feature is character consistency, which allows users to edit or generate images of people, pets, or objects while retaining their recognisable characteristics across multiple changes.

The tool also supports multi-image fusion, enabling the blending of several images into a single, coherent composition.

Users can employ multi-turn editing, where they progressively refine an image with a series of commands, such as changing backgrounds, adding new elements, or altering styles. Another feature, design mixing, lets users transfer stylistic aspects such as colour, texture, or patterns from one image to another.

Safety and watermarking measures

To address concerns about the authenticity of AI-generated content, Google has incorporated visible watermarks alongside its proprietary SynthID invisible watermarking technology. These measures are aimed at ensuring transparency and reducing the misuse of the tool for creating misleading or harmful imagery.

The company has also implemented safeguards within the model to prevent the generation of inappropriate or non-consensual content. These guardrails align with Google’s broader efforts to promote responsible use of AI systems.

Availability and developer access

Nano Banana is available to both free and premium users of the Gemini app on Android and iOS. Developers can also access the model through the Gemini API, Google AI Studio, and Vertex AI, allowing integration into a wider range of applications and enterprise workflows.