Google's image generation AI, 'Nano Banana 2' and 'Nano Banana Pro,' are finally available to the general public; they are now available via API after the preview period ends.

Google has announced the official release of its image generation AI, '

🍌 Nano Banana Pro [gemini-3-pro-image] and Nano Banana 2 [gemini-3.1-flash-image] are now GA and ready for production via the Gemini API. Check out these great community examples to see the capabilities of both models in action 🧵↓ pic.twitter.com/Gvv7AFwHu7

— Google AI Developers (@googleaidevs) May 28, 2026

Nano Banana is the nickname for 'Gemini 2.5 Flash Image,' an image generation AI announced by Google in August 2025. In November 2025, 'Gemini 3 Pro Image (Nano Banana Pro),' built on Gemini 3 Pro, was announced. Nano Banana Pro offers superior information visualization compared to Nano Banana, enabling it to output images that better match the atmosphere of the given information, and making it easier to generate and edit image content containing text, such as infographics and Japanese text, without distortion.

Google's image generation AI 'Gemini 3 Pro Image (Nano Banana Pro)' has been released, leveraging Gemini's inference capabilities to enhance language support and information organization - GIGAZINE

Furthermore, in February 2026, 'Nano Banana 2 (Gemini 3.1 Flash Image)' was released, which maintains the same quality as Nano Banana Pro while offering faster processing. You can find more information about the features of Nano Banana 2 in the following article.

Google releases image generation AI 'Nano Banana 2,' enabling faster and more accurate image generation - GIGAZINE

Nano Banana Pro and Nano Banana 2 were previously available as preview versions in AI Studio and Gemini API, but the general availability (GA) version for general developers was released on May 29, 2026.

The Gemini API page provides sample code for using the Gemini API programmatically, covering image generation and editing with Nano Banana, as well as new features in Nano Banana Pro and Nano Banana 2, such as generating images by combining up to 14 reference images, generating images based on real-time information from Google Search, generating images from videos, and generating images with up to 4K resolution.

Another key feature of Nano Banana Pro and Nano Banana 2 is that they are thinking models that use inference processes to handle complex prompts. They can generate up to two intermediate images to test the construction and logic, and you can visually see the inference process leading up to the final image.

For security reasons, all images generated with Nano Banana Pro and Nano Banana 2 have a SynthID watermark embedded in them. Additionally, the image generation function based on real-time information from Google Search cannot use real-life images of people obtained from web searches.

In addition, coinciding with the official release of Nano Banana Pro and Nano Banana 2, we have added support for video files in input prompts as a preview of Nano Banana 2's new features. By analyzing the visual context, specific subjects, and actions within the video footage, you can generate context-aware images such as thumbnails and infographics.

Some features of the Nano Banana Pro and Nano Banana 2 APIs are available in the free tier. You can see the cost per token on the following page.

Gemini Developer API Pricing | Gemini API | Google AI for Developers
https://ai.google.dev/gemini-api/docs/pricing?hl=ja

Additionally, Google AI Studio allows you to use Nano Banana Pro and Nano Banana 2 by upgrading to a paid plan. Nano Banana is available for free.

Google Accounts
https://aistudio.google.com/