Nano Banana Pro: How to Use Google’s Gemini 3 Pro-powered Image Tool and What It Can Do

Google has released Nano Banana Pro, an upgraded AI-powered image creation and editing tool available through the Gemini app and website. Also called Gemini 3 Pro Image, this tool offers advanced capabilities such as improved text rendering, higher-quality outputs, better contextual understanding, and support for more reference images.

Many users are unsure where Nano Banana Pro is located or how to use it. Here is a clear explanation.

How to Use Nano Banana Pro

Nano Banana Pro is accessible inside the Gemini mobile app and through Google’s developer platforms.

Steps to use it in the Gemini app:

1.Open the Gemini app on Android or iOS, or visit the Gemini website.

2.Click “Create images” from the home screen options.

3.If you don’t see it, open the Tools menu in the text box and select it.

4.From the model menu at the bottom right, choose "Thinking" (this switches to Nano Banana Pro).

5.Type a clear prompt describing subject, action, style, camera angle, lighting, mood, aspect ratio, or resolution.

6.For editing an existing image, upload one or more reference images and specify what to change.

7.If you are not satisfied with the results, send follow-up prompts to refine the output.

8.Once final, tap the image and download it.

    Free and Pro users will see visible watermarks.

    Ultra users can remove visible watermarks (invisible SynthID remains).

What Is Nano Banana Pro?

Nano Banana Pro is Google’s advanced image generation tool built on Gemini 3 Pro, replacing the earlier version powered by Gemini 2.5 Pro. It specializes in detailed edits and high-quality image generation while preserving consistency and accuracy.

What’s New in Nano Banana Pro?

Combine up to 14 reference images in one prompt.

Maintain consistent appearance for up to 5 people across multiple outputs.

Highly precise editing—adjust lighting, focus, camera angle, or convert day-to-night.

Supports 2K and 4K high-resolution image output.

Improved text rendering with support for long text, multiple fonts, and more languages.

All images include SynthID invisible watermark, and visible watermark for non-Ultra users.