This year, Google’s AI image generation team achieved a feat unmatched by its competitors. The Nano Banana image generation model (Gemini 2.5 Flash Image) went viral not once, but twice during pre-release testing online. In August, Nano Banana initially made waves, though it was just a placeholder name until Google officially adopted it. Then, a few weeks ago, Nano Banana 2 (also called Nano Banana Pro) captivated testers with significant improvements surpassing the original’s capabilities. Rumors suggested that Google would launch Gemini 3 Pro and Nano Banana 2 simultaneously, but that was not the case. Gemini 3 Pro debuted on Tuesday, followed by the release of Nano Banana Pro (Gemini 3 Pro Image), Google’s latest cutting-edge model, on Thursday.
Notably, Google also introduced a new Generative UI experience for Gemini 3, where the AI crafts immersive visuals tailored to prompts requiring rich visual interpretation. Although Google did not disclose the specific image generator powering this experience, the launch of Nano Banana Pro fits perfectly within this context. Built on Gemini Pro technology, Nano Banana Pro leverages state-of-the-art reasoning and real-world knowledge to deliver superior visualization of information.
What’s new in Nano Banana Pro?
Google’s Thursday announcement confirmed many of the features that had leaked last week. The new model offers enhanced character consistency and improved text rendering capabilities over its predecessor, which already produced realistic images with text support. Nano Banana Pro excels at blending elements seamlessly; it can incorporate up to 14 images in one output while maintaining character consistency for up to five people.
The model also introduces localized editing features, allowing users to adjust focus, apply sophisticated color grading, modify lighting, and add bokeh effects to images. Another significant upgrade is the resolution capability—users can now generate images in 2K and 4K, a marked improvement over the original Nano Banana’s 1024 x 1024 resolution. Additionally, Nano Banana Pro supports a wider range of aspect ratios, offering greater flexibility for creators.
Images generated by Nano Banana continue to be protected by an invisible SynthID digital watermark. The Gemini app can identify images created with Google’s AI generators. For Gemini Free and Gemini AI Pro users, generated images will display a visible watermark (the Gemini sparkle) to clearly indicate their origin, while Ultra plan subscribers enjoy watermark-free images.
Where is Nano Banana Pro available?
Nano Banana Pro is primarily accessible through the Gemini app, available under the Thinking model when users select the “Create Images” option. Free Gemini plan users receive a limited quota of Nano Banana Pro images before reverting to the original Nano Banana model, while AI Plus, Pro, and Ultra subscribers benefit from higher quotas. U.S. users can also access Nano Banana Pro via AI Mode. Additionally, NotebookLM users worldwide have access to these new models, and Google AI Ultra subscribers can utilize Nano Banana Pro within the Flow AI filmmaking tool.
Google also announced plans to integrate Nano Banana Pro into Google Ads, Workspaces, Google Slides, and Vids. Developers can harness the model in Google AI Studio, Google Antigravity, and through the Gemini API. While Google hinted at future availability through third-party services supporting its AI tools, no specific partnerships were mentioned in the announcement.
