Google launches ‘Nano Banana’ for next-gen image editing; check key features

Tech giant Google has finally rolled out its new image generation and editing AI model, Gemini 2.5 Flash Image, also known as Nano Banana. The model edits images with great detail, and can be used for multiple things, including blurring images, removing objects, and adding colours to a black and white photo, among other things.

The model has a ‘deep, semantic understanding of the real world’, making it more advanced than other image generation models, Google claims.

Google CEO Sundar Pichai made waves on the social media platform X after he randomly posted three bananas, giving rise to speculations that Google has launched a new AI tool with some relation to bananas.

Interestingly, after an X user Nimisha Chanda asked what this tweet meant, X’s AI platform Grok replied, “Sundar Pichai's [banana emojis] likely teases Google's rumoured 'Nano Banana' AI tool for precise image editing and generation."

The rumours of Google being behind the new AI model have been doing the rounds for a couple of weeks, as the company started teasing by posting random and vague tweets.

Soon after Pichai posted the mysterious tweet with three bananas, he confirmed that Google is rolling out its new image editing model in the Gemini app.

How does Nano Banana work?

Google has shared blog posts on how Nano Banana works, its features and how users can do advanced editing through the model.

The model allows you to blend several images into a single image, maintain character consistency for enhanced storytelling and make targeted transformations using natural language. Through the model, users can also change location or clothes in the image, do multi-turn editing and mix designs.

Nano Banana is currently available through the Gemini API and Google AI Studio for developers and Vertex AI for enterprise.

The model is priced at around ₹2,496 ($30) per 1 million output tokens. This makes each image priced at ₹3.24 ($0.039 per image).

Many reports are already suggesting that the model is extremely advanced and has the potential to replace Photoshop.

A Reddit user, writing about his experience, titled his post ‘Google really raised the bar with nano banana, scary how good and accurate it is.’ Others commented on the post, discussing how the model can see and notice things that might not be visible to the naked eye.