Late Thursday night Beijing time, the Nano Banana image generation model under the AI giant Google received a major update – Nano Banana Pro made a dazzling debut!
As a background, Google released the Nano Banana based on the Gemini 2.5 Flash model at the end of August this year. Because this model can turn characters or anime photos into lifelike “3D printed figurines”, it has created a global breakout effect in a short time. Google CEO Pichai disclosed that driven by Nano Banana, the monthly active users of the Gemini app soared from 450 million to 650 million in just one quarter.
With the release of Google’s new flagship large model this week, Google’s image generation model has also witnessed an expected leapfrog upgrade.
In Google’s words, Nano Banana Pro can “turn users’ ideas into studio-level designs with unprecedented control, flawless text presentation and enhanced world knowledge”.
According to Google, the Nano Banana Pro is based on the Gemini 3 Pro released earlier this week. It not only outperforms the previous generation Nano Banana in terms of detail expressiveness, image resolution and text rendering accuracy, but also can generate text in different styles, fonts and languages. The new generation of image generation models has also added editing functions and incorporated Internet search capabilities.
Google emphasizes that Nano Banana Pro is the best AI model for generating correct and clearly readable text in images, whether it is a short slogan or a long paragraph, and supports richer textures, fonts and calligraphy. At the same time, with the enhanced multilingual reasoning capabilities of Gemini, users can generate text in multiple languages or translate and localize content.