Last year in August, Google announced Nano Banana (Gemini 2.5 Flash Image), an image generation and editing model. The model was state-of-the-art in several key AI image generation benchmarks at that time, and it went viral among consumers for image editing. Later in December, Google announced Gemini 3 Pro Image (codenamed Nano Banana Pro), taking advantage of Gemini’s state-of-the-art reasoning and real-world knowledge to generate more accurate images.
Today, Google announced Nano Banana 2 (Gemini 3.1 Flash Image). Google claims that this model is a state-of-the-art image model which can deliver Nano Banana Pro-level images at the speed of Nano Banana. Also, this new affordable image model supports the following features, which were limited to the Pro model previously:
- Real-world knowledge: Nano Banana 2 can now take advantage of Gemini’s real-world knowledge base, which includes real-time information and images from web search. Google highlighted that this model can now create even better infographics and visualizations.
- Legible text: Nano Banana 2 can now generate accurate, legible text, which has been quite a challenge for all leading image generation models. Google will even allow users to translate and localize text within an image.
Apart from the above, Nano Banana 2 comes with several improvements over Nano Banana. For example, this new model can now maintain character resemblance for up to five characters and the fidelity of up to 14 objects in a single workflow. Also, the model can now more strictly follow complex requests.
In Nano Banana 2 you can generate images with new aspect ratios: 4:1, 1:4, 8:1 and 1:8.
— Nano Banana 2 (@NanoBanana) February 26, 2026
Try it with these prompts:
1. A 4:1 panoramic photo I took on my phone, showing the luscious jungle view from my cabin
2. A 4:1 panoramic photo I took on my phone, showing the view from my… pic.twitter.com/U1IyaQWNkT
This model now supports the generation of images in various aspect ratios and resolutions from 512px to 4K, making it suitable for a wide variety of scenarios. Finally, Nano Banana 2 delivers vibrant lighting, richer textures, and sharper details when compared to its predecessor.
Nano Banana 2 is now available for consumers, developers, and enterprises. For consumers, Nano Banana 2 will replace Nano Banana Pro as the default image generator across the Fast, Thinking, and Pro models in the Gemini app. This new model will also be available in Google Search AI Mode and Google Lens. For developers, Nano Banana 2 is available via the Gemini API in AI Studio and Google Cloud.