Google has introduced Nano Banana 2, its latest AI image-generation model, also known as Gemini 3.1 Flash Image. The new system arrives shortly after the release of Gemini 3.1 Pro and is set to replace Nano Banana Pro across the Fast, Thinking and Pro model tiers within the Gemini ecosystem.
While Nano Banana Pro will remain available for specialized workflows, access will be limited to Google AI Pro and Ultra subscribers through the platform’s advanced settings menu. Nano Banana 2 is also expanding beyond Gemini, with integration planned for AI Mode in Google Search. Users will be able to access the model through the Google app as well as standard mobile and desktop web browsers.
According to Google, Nano Banana 2 combines the advanced reasoning and world knowledge of its predecessor with significantly faster processing speeds. The model is designed for rapid visual editing and iterative creative workflows, enabling users to refine images quickly without sacrificing quality. Its capabilities are enhanced through real-time information and imagery retrieved from web search, allowing more accurate rendering of specific subjects and contexts.
The model also introduces improved comprehension of structured content. It can transform written notes into diagrams, generate infographics and create data visualizations with enhanced clarity. Text generation within images has also been refined, producing more accurate and legible results. Additionally, Nano Banana 2 can translate and localize text directly inside generated visuals, expanding its usefulness for global audiences.
Google highlights major improvements in visual fidelity and consistency. The system can maintain recognizable character features across up to five subjects within a single project while preserving detail across as many as 14 objects in one workflow. Enhanced lighting, richer textures and sharper details aim to deliver more photorealistic imagery across a wide range of prompts.
By narrowing the gap between speed and visual quality, Nano Banana 2 represents a significant step forward in AI-powered image creation. The rollout positions the model as a core component of Google’s evolving generative AI strategy across search, productivity tools and creative applications.