Google's Nano Banana Makes the Internet Go Bananas
Google has confirmed its Nano Banana image-editing tool, which recently gained popularity, is now integrated into the Gemini app. The company stated on Tuesday that this project, known for its ability to generate and refine images, addresses a previous challenge for artificial intelligence makers.
The model, now officially named Gemini 2.5 Flash Image, became available on Tuesday for both free and paid Gemini users via the web and mobile applications. Free users can perform up to 100 image edits daily, while paid subscribers have access to ten times that amount.
Previously, under its codename Nano Banana, the model garnered significant social media attention and outperformed other models on the LMArena charts for image editing. This demonstrated its advanced capabilities even before its official unveiling.
The new tool allows users to create photos from text descriptions or by modifying existing images. Google provided examples, including transforming a woman into matador attire, and imagining multi-stage room renovations.
The Gemini 2.5 Flash Image also supports combining two distinct images, such as applying a transformed pattern from one photo onto boots in another. Google reported the model is more effective at complex, multi-step editing tasks.
A key advancement is the tool’s ability to maintain a subject’s identity through various edits, preventing them from appearing like someone else. Google noted this is crucial when modifying photos of individuals or pets.
All creations generated by the model are distinctly labelled as AI-generated, utilising Google’s SynthID watermarking system.
While artificial intelligence has proven adept at image generation, editing, particularly with real individuals, has presented difficulties for developers. This refinement capability marks a notable step in AI image technology.
The company confirmed its goal is for friends, family, and pets to consistently resemble themselves in edited photos, from a 1960s beehive haircut to adding a tutu on a chihuahua.
This development follows OpenAI’s significant increase in ChatGPT downloads and usage after releasing its highly capable image generator in March. The ongoing advancements in AI tools also raise concerns about the potential for generating deepfakes and misinformation.
Google confirmed its Nano Banana image-editing tool is now integrated into the Gemini app.
The tool, officially named Gemini 2.5 Flash Image, offers advanced image generation and refinement capabilities.
It maintains subject identity during edits and is available to free and paid Gemini users.
Source: AXIOS
