Google is upgrading its image generation model with new editing features, higher resolution, more accurate text rendering, and web search capabilities.
The new model, called Nano Banana Pro, is built on Google’s latest large-scale language model, Gemini 3, which was released earlier this week. According to the company, Nano Banana Pro improves on the previous generation Nano Banana with the ability to create more detailed images and accurate text, and generate text in a variety of styles, fonts, and languages.
This model also has a web search feature, so you can search for recipes, generate flashcards, and more.
Google says Nano Banana Pro is aimed at giving professionals more control over their images, allowing users to control aspects such as camera angle, scene lighting, depth of field, focus, and color grading. Users can also generate 2K or 4K images with Nano Banana Pro, compared to Nano Banana’s 1024 x 1024px resolution limit.
The company noted that while the Nano Banana Pro can produce high-quality images, it is slower and more expensive than the original model, which costs $0.039 per 1024-pixel image. In comparison, the new model costs $0.139 per 1080p or 2K image and $0.24 per 4K image.
The new model allows you to use 6 high-fidelity shots and blend up to 14 objects in an image. You can also maintain consistency and similarity between up to five people. The company has released a demo app where you can try out some of these features.
Nano Banana Pro is deployed in many of Google’s existing AI tools. The Gemini app will now use the new model by default to generate images, but users on the free subscription tier will be able to generate a limited number of images using this model, after which it will default to the original Nano Banana model.
tech crunch event
san francisco
|
October 13-15, 2026
Higher generation thresholds apply to Google AI Plus, Pro, and Ultra subscribers, but the company hasn’t disclosed the exact limits. These subscribers also have access to models within NotebookLM.
Google is also making this model available for search via AI mode for AI Pro and Ultra subscribers in the US. Ultra subscribers can access the model in Flow, the company’s video tool, and it’s also available for Workspace customers in Google Slides and Vids.
Developers can leverage Nano Banana Pro through the Gemini API, Google AI Studio, and the company’s new IDE Antigravity.
The company is also incorporating SynthID, a technology that watermarks and detects AI-generated images, into its Gemini app. When a user uploads an image, the chatbot tells them if the image was created or modified by the company’s image model.
Google said that over time, support for C2PA content credential detection for content validation will be included.
This story has been updated to reflect Google’s approach to supporting C2PA.
