Google said on the 20th (local time) that it will release Nano Banana Pro (Gemini 3 Pro image), which further enhances the capabilities of Nano Banana.
While the existing Nano Banana was based on Google's AI model Gemini 2.5, Nano Banana Pro is a model built using Gemini 3 Pro, unveiled on the 18th. Accordingly, Nano Banana Pro is equipped with the ability to visualize and design ideas based on the latest reasoning capabilities, and it is also linked to Google Search to reflect the latest information in real time. For example, Google said that even if a user simply enters the command (prompt) "Draw an infographic showing how to make cardamom black tea," it generates fact-based illustrations.
Unlike conventional image-generation models, Nano Banana already handled lettering accurately, and this has been further improved in Nano Banana Pro. Even when calligraphy with design elements or images with integrated text are generated, the fonts and textures are created naturally. In particular, by leveraging Gemini's multilingual reasoning, its ability to render multiple languages, including Korean, has been enhanced.
Users can input up to 14 images and use them as components to create a new image. In this process, the images initially entered maintain consistency and are applied naturally to the generated output. It also has editing functions such as changing the angle or focus of an existing image, or switching the lighting to make a daytime photo look like a nighttime photo. The generated images support various aspect ratios, and the resolution can be set up to ultra-high-definition 4K.
General users (including students) can use Nano Banana Pro starting today by selecting Image generation under the Inference model in the Google Gemini app. However, usage may be limited for free users. If free usage is exceeded, it switches to the existing Nano Banana. Nano Banana Pro is also available in the NotebookLM work assistant, and for paid Google subscribers in the United States, it is provided in the AI mode of the Google Search bar. Developers and professionals can use Nano Banana Pro through dedicated tools such as Workspace, AI Studio, API, and Flow.
Google added a verification feature to the Gemini chatbot so users can check whether an image was generated by AI. Media generated with Google tools contains an invisible digital watermark, "SynthID," and Google said that if a user simply uploads an image to the Gemini app and asks, "Was this generated by AI?" the app verifies and reports the result. Google will first apply this feature to images and then expand it to audio and video.
In addition, images generated by free users and paying users on the Pro plan will retain a visible watermark. However, users on the Ultra plan for professionals can generate images without a visible watermark.