Nano Banana Pro Preview

Nano Banana Pro is built on Gemini 3 Pro as Google's most advanced image-generation and editing model, representing a significant step beyond the original Nano Banana. The model prioritizes multimodal reasoning and real-world grounding alongside high-fidelity visual synthesis, enabling it to produce context-rich graphics ranging from infographics and diagrams to cinematic composites. Its industry-leading text rendering supports long passages and multilingual layouts within images, while consistent multi-image blending and accurate identity preservation across up to five subjects make it capable of handling complex multi-element compositions. Fine-grained creative controls let users apply localized edits, adjust lighting and focus, transform camera angles, and output in 2K/4K with flexible aspect ratios.

The model scored 1,485 on the Arena Leaderboard, establishing a competitive performance reference. It supports Search grounding to pull real-time information into visual outputs, which helps with design work that demands current data integration. Targeted use cases include professional-grade design, product visualization, storyboarding, and complex creative workflows. Notably, tool-use capability is not currently tracked, which means it may not fit agentic or function-calling patterns out of the box. For workflows requiring image or audio processing beyond direct input, a separate model in the pipeline is needed.