The latest and most powerful AI model from Google, Gemini 3, is now available at gemini.google.com. Select “Thinking” to use this new model. Gemini 3 is topping AI leaderboards and benchmarks across almost all capabilities. Gemini 3 is particularly notable for it native "multimodality", i.e. its ability to understand and reason over text, images, video and audio.
Google also released Gemini 3 Pro Image, also known as "Nano Banana Pro”. This is an advanced AI model designed for creating and editing images with studio-quality precision and control. Capable of generating sharp, legible text for posters and diagrams, the model supports complex creative workflows such as prototyping and design translation for different locales. It leverages Gemini's extensive real-world knowledge and deep reasoning to produce highly accurate results, ranging from detailed infographics to historically faithful scenes. To use the pro version, go to gemini.google.com, select “Thinking” for the model and click the “Create Image” button.