Google DeepMind has announced the introduction of "Nano Banana Pro," a new image generation and editing model. Based on Gemini 3 Pro, this model leverages advanced reasoning capabilities and real-world knowledge. The company stated that it enables the generation of more accurate and contextually rich visuals compared to the previous "Nano Banana."
NewsGoogle DeepMindGemini 3 ProNano Banana Pro
Google DeepMind Releases Nano Banana Pro, an Image Generation Model Based on Gemini 3 Pro
This article is a translation. Read the Japanese original
Nano Banana Pro is intended for use in prototyping, transforming data into information graphs, and converting handwritten notes into diagrams. Through the reasoning power of Gemini 3, it supports the creation of more practical content, such as educational explanations and infographics. Additionally, it can connect to the Google Search knowledge base to visualize real-time information, such as recipes and weather.
Text rendering precision has also been significantly improved. The company stated that the model can accurately display legible text in multiple languages directly within images. It can express everything from short slogans to long paragraphs using a variety of fonts and typefaces. This is expected to facilitate the creation of mockups and posters, as well as international content localization.
The model is currently available through Google products such as the Gemini app, Google Ads, and Google AI Studio. As the successor to the Gemini 2.5 Flash Image model "Nano Banana" released several months ago, it aims to further enhance the expressive power of creators.
Source: Nano Banana Pro (HN 1275pt, 685 comments) (HN Search (backfill), 2025-11-21)