ChatGPT had had its viral moment with the Ghibli trend, and Google seems to having its own with Nano Banana.
The new image editing Nano Banana model has brought 10 million new users to Gemini, a Google exec has said. Josh Woodward, who is a VP at Google, said that Nano Banana has edited 200 million images so far. It’s also brought 10 million new users to Google’s Gemini platform. “TPUs red hot, SRE pagers howling,” he posted on X, suggesting that Google was running at peak capacity with all the new demand.
This trend seemed to be corroborated by SimilarWeb data. SimilarWeb shared a graph of traffic to Google AI Studio, which showed that the daily web visits to Google AI Studio had jumped from around 3 milllion a day to around 4.5 million a day right after Nano Banana was released on 26th August. Google AI Studio’s traffic has been steadily increasing over the past year — it saw around 300k daily users in September 2024, but that has risen 15x to 4.5 million users now.
Nano Banana had first appeared as a model on LMArena, a platform where users do blind tests of outputs of two different AI models. Nano Banana quickly created plenty of buzz for its realistic image generation capabilities and its image editing features. Google began hinting that it was behind the model, with Google employees sharing images of a single banana emoji on X. Google DeepMind CEO Demis Hassabis and other senior leaders too teased the launch of the model with banana references on X. Google finally unveiled Nano Banana as Gemini Flash 2.5 Image, and made it available through Gemini, Google Image Studio, and its API. Google has even created an X account that automatically generates images to get more people to use it.
Images have ended up being a popular way for AI companies to go viral. Stability AI’s Stable Diffusion image generation model had thrust the company into mainstream limelight in 2023. Earlier this year, OpenAI’s image generation model went mega viral with the Ghibli trend, and the company saw its userbase expand massively because of that one feature. Elon Musk has been aggressively pushing Grok’s Imagine video-generation capabilties, hoping it would catch on among users. Google has now created an image generation model that seems to be much better than the competition, and going by the numbers, seems to be bringing millions of new people to Google’s AI offerings.