Google first launched the Nano Banana AI image generator in August last year. The tool, also called the Gemini 2.5 Flash Image AI model, focused on faster image creation. It also improved element-based editing. In addition, it delivered better character consistency.
Later, in November 2025, the company introduced Nano Banana Pro powered by Gemini 3 Pro. At that time, it supported image generation up to 4K resolution. Moreover, Google described it as an advanced image editing and creation system.
Now, the Mountain View-based company has launched Nano Banana 2. This new version aims to provide Pro-level image quality. At the same time, it offers noticeably faster performance.
Availability and Supported Platforms
According to a recent blog post, Nano Banana 2 has started rolling out. Users can access the Gemini 3.1 Flash Image AI model across multiple Google services.
Inside the Gemini app, Nano Banana 2 replaces Nano Banana Pro in the Fast, Thinking, and Pro AI modes. However, Google AI Pro and Ultra subscribers can still regenerate images with the older model for specialised work.
Additionally, the tool is available in 141 countries, including India. It also supports eight new languages. Users can try it in Search AI Mode and Google Lens through the Google app. Meanwhile, it also works on both mobile and desktop browsers.
Furthermore, Google released Nano Banana 2 as a preview in AI Studio, the Gemini API, and Vertex AI. It has now become the default image generator in Flow, Google’s video creation platform. As a result, Flow users can use it without spending credits.
The company has also integrated the model into Google Ads. Therefore, advertisers can create banners, campaign creatives, and ad ideas more easily.
Key Features and Improvements
Nano Banana 2 uses Gemini’s advanced knowledge base. Because of this, it can pull real-time information and reference images from web search. Consequently, the system produces faster and more accurate subject-focused images.
Moreover, the tool can generate infographics, diagrams, and data visualisations in greater detail. It can also create clear and readable text inside images. Therefore, it becomes useful for marketing materials and greeting cards.
In addition, the system can translate and adapt text that already exists inside an image.
Image Quality and Performance
Google claims Nano Banana 2 produces high-quality photorealistic images quickly. For example, it can maintain consistency for up to five characters in one workflow. Likewise, it can preserve visual accuracy for up to 14 objects.
The system now follows instructions more precisely. As a result, it understands complex prompts and subtle details much better.
Finally, like Nano Banana Pro, the model supports 4K resolution image generation. It also delivers brighter lighting, richer textures, and sharper details.