Google Unveils Next-Gen Nano Banana 2 AI Image Model for Gemini

Google has unveiled the next iteration of its popular Nano Banana AI image generation model, called Nano Banana 2. The new release, also known as Gemini 3.1 Flash Image, promises to deliver results similar to the Nano Banana Pro model but with the speed of the non-pro Flash variant.

According to Google, the Nano Banana 2 model has been enhanced with more advanced world knowledge, drawn from the broader Internet by the Gemini 3.1 Large Language Model (LLM). This expanded knowledge base allows the AI to render objects with greater fidelity and create more accurate infographics. Google also claims the new model has achieved Pro-like text accuracy in its image outputs, putting an end to the days of squiggly AI-generated text.

The release of Nano Banana 2 is the latest in a series of rapid-fire AI model updates from Google, as the tech giant works to keep pace with rivals like OpenAI and Anthropic. The original Nano Banana model debuted last year and has since received several updates, each one improving on the capabilities of the previous iteration.

With the Nano Banana 2 release, Google is positioning its image generation technology as a strong competitor in the increasingly crowded AI art and content creation space. The new model's ability to generate high-quality, accurate imagery at speed is likely to appeal to a wide range of users, from graphic designers to social media marketers.

The Nano Banana 2 model is available starting today as part of the Gemini 3.1 update. Users of the Gemini app can now access the new image generation capabilities and begin creating their own unique, AI-powered visuals.