Google DeepMind has introduced Nano Banana 2, the latest version of its image generation model, as it works to make AI-powered creativity faster and more dependable.

The new model blends the quick response time of Gemini Flash with the more advanced visual reasoning that users previously saw in Nano Banana Pro. In simple terms, it is designed to deliver high-quality images without making users wait.

Also known as Gemini 3.1 Flash Image, Nano Banana 2 is being rolled out across a wide range of Google platforms, including the Gemini app, Search, Google Cloud, AI Studio, Flow and Google Ads. By integrating the model across its ecosystem, Google is making image generation a built-in feature rather than a separate tool.

The company says the updated model has a better understanding of prompts and produces results that are more accurate and grounded. It combines Gemini’s broader knowledge base with real-time information from web search, helping ensure that generated visuals reflect relevant and up-to-date context.

For users, this means the ability to create infographics, turn rough notes into clear diagrams, and build data visualisations with improved accuracy. The model also handles text within images more clearly and supports in-image translation, which could be useful for businesses creating marketing materials for different regions.

Nano Banana 2 also focuses on visual consistency. It can maintain character resemblance for up to five people and preserve detail across as many as 14 objects in a single workflow. This makes it easier to develop storyboards or structured visual narratives without repeatedly refining prompts.

Google notes that the system follows detailed instructions more reliably and supports various aspect ratios and resolutions, from 512 pixels to 4K. While Nano Banana Pro will remain available for tasks requiring the highest level of factual precision, Nano Banana 2 is positioned as the faster, more flexible option for everyday creative work.

Within the Gemini app, Nano Banana 2 will replace Nano Banana Pro across Fast, Thinking and Pro modes. However, Google AI Pro and Ultra subscribers will still be able to regenerate outputs using Nano Banana Pro when needed.

The rollout also includes AI Mode and Lens in Search on both mobile and desktop, expanding to 141 additional countries and eight more languages. Developers can access the model in preview through AI Studio and the Gemini API, including via Vertex AI on Google Cloud. In Flow, it becomes the default image generation model, and it will power image suggestions during campaign creation in Google Ads.

Nano Banana first launched in August 2025, followed by Nano Banana Pro in November. With Nano Banana 2, Google appears to be streamlining its image tools into one faster, more widely available model aimed at meeting everyday creative demands.

Also Read: Google Brings AI-Powered Music Creation to Gemini With Lyria 3