Google has been constantly improving its image generation model to stay on top of rivals. When Google first launched Nano Banana, it took the internet by storm. The internet was flooded with images generated using Nano Banana. Well, Google has now upgraded its image generation model, which means we now have Nano Banana 2. The new image generation model, which is technically Gemini 3.1 Flash Image, can create more realistic images than its predecessor. Google says that the new image model will be the default model in the Gemini app for its Fast, Thinking, and Pro modes.
Nano Banana 2 image generation model brings faster image generation in multiple aspect ratios
The new Nano Banana 2 image model retains some of the high-fidelity characteristics of the Nano Banana Pro model that was launched back in November 2025. However, the new model can create images faster. Interestingly, Nano Banana 2 allows you to create images with a resolution ranging from 512px to 4K, and that too in different aspect ratios.
If you are looking for character consistency, Nano Banana 2 should be your go-to choice. Google, in the announcement blog post, mentions that it can maintain character consistency for up to five characters and fidelity for up to 14 objects in one workflow for better storytelling. If you want detailed images, the good news is that it also handles complex requests. Elements including vibrant lighting, richer textures, and sharper detail are all supported by the Nano Banana 2 image generation model.
Notably, the new model will also be the default for anyone using Google’s video editing tool, Flow. For those looking to generate images directly from Search, Nano Banana 2 will become the default for Google Search results via Google Lens and in AI Mode across 141 countries on the Google app. It will also be available on the web across desktop and mobile.
All the generated images features SynthID
If you want to continue using Nano Banana Pro for custom tasks by regenerating images via the three-dot menu, you must be subscribed to the Google AI Pro or Ultra plan. Not to forget, Nano Banana 2 will be available in preview for developers via the Gemini API, Gemini CLI, and the Vertex API. It will also be available through AI Studio and Antigravity, which Google launched late last year.
Last but not least, Google has also confirmed that all images generated using Nano Banana 2 will feature SynthID. To catch you up, it is Google’s in-house watermarking tool that allows detection of AI-generated images.
Have you tried the new image generation model yet? If yes, let us know the improvements you experienced compared to its predecessor.
