Google launched Nano Banana AI Photo Editor, Sundar Pichai Goose Banana with her dog photo
For weeks, Nano Banana has created a discussion among beginners and AI enthusiasts due to the ability to edit photos. Now, Google revealed that it is actually a company behind this AI model and Banana is now rolling for Gemini users.
Yesterday evening, Google CEO Sundar Pichai tweeted three banana emozis, confirming that his company is actually behind the viral sensation Nano banana. Hours later Google officially introduced the AI tool, which aims to allow users to edit and modify photos with a simple prompt. Although the code named Nano Banana, the new tool has been introduced as Mithun 2.5 flash image.
The generic image AI model is now integrated in the Gemini app and promises high quality, reference-covered image generation abilities, keeping the original subjects consistent.
Showing the capabilities of this AI new image editor, Google CEO Sundar Pichai took X (East Twitter) and shared a series of fickle images of his dog Jeffrey, edited with new tools in the celebration of International Dog Day. In the pictures, Pichai re -prepared the Jeffrey Surfing Waves, donated a shepherd cap, twisted the superhero, and even dressing as a chef. The new AI model has edited dog photographs with different subjects, while all have their looks alike.
“Our image editing model is now rolling in @Geminiapp – and yes, it is (banana emozis). Edit the leaderboard at the top of the image of @lmarena, it is especially good in maintaining equality in various references. See some of my dogs Jeffrey in honor of the day of the international dog – though not telling you the fools,”
How Gemini 2.5 flash image model works
The new image editor of Google makes the first image generation characteristics in the Gemini app, which the company introduced earlier this year. According to Google, AI image is one of the most common frustrations with generation when an AQI edited photo of a person or pet does not look like them. But the purpose of nano banana update is to solve that problem.
To use this new model, users simply need to upload a photo, provide a prompt, and the AI of Gemini can apply a wide range of changes, recognizing the subject.
With simple prompt-based editing, the new AI image editing model also introduces multi-turn editing, making users step by step. Another feature design is mixing, which allows users to borrow patterns and textured from one photo and apply them to another. For example, you can imagine the butterfly feathers to convert a cloth print for a dress or give rain shoes.
Additionally, the new image generation model supports combining multiple photos in a spontaneous image. Users can also feed their edited images back to Gemini to generate animated videos.
Nano banana is available globally
Google is rolling the Gemini 2.5 flash image model for free and paid users in Genini App in Geninia App today. To ensure transparency, each AI-borne image will take a visible watermark with the invisible syntid digital watermark of Google, making it clear that the pictures are AI-Janit.
For developers and enterprises, Google is also providing the new Gemini 2.5 flash image model through Gemini API, Google AI Studio and Wartax AI. The pricing is set on $ 30 per million output tokens, with a cost of $ 0.039 to generate each image.