How to use awesome-nanobanana-pro: prompts, examples and tips

What awesome-nanobanana-pro is and what it is not

The repository is a text list, not software. The README describes it as a curated collection of Nano Banana prompts, image generation styles, and resources. The prompts are gathered from X (Twitter), WeChat, Replicate, and prompt engineers, and each one usually credits its source.

You do not install anything to use it. You need your own access to the Nano Banana Pro image model, for example through Gemini or another platform. The list does not provide that access.

• Good for: copy-ready prompts, visual style ideas, learning how prompt wording affects output
• Not for: installing a tool, verifying that outputs are reproducible, or formal documentation of the model

Open the repository and find the section you need

Start from the table of contents in the README. The prompts are sorted into themed sections, including Photorealism & Aesthetics, Creative Experiments, E-commerce & Virtual Studio, Interior Design, Social Media & Marketing, and Photo Editing & Restoration. Each prompt has a short description of what it is designed to test or produce.

If you want a local copy to browse offline, the README gives this optional command:

• Open the README on GitHub and use the table of contents links
• Pick a section that matches your goal, such as product images or a specific era
• Read the description line under each heading to see what the prompt is for

Copy a prompt and adapt it to your own subject

Each entry has a Prompt block you can copy directly. Many prompts describe a subject, camera style, lighting, background and mood in separate parts. Changing one part at a time lets you see how wording affects the result.

Some prompts refer to a reference image, for example to keep a face accurate. Those only work if your model setup accepts an uploaded image. Some prompts are written as JSON; you can paste them as they are or simplify them into plain sentences.

• Copy the prompt text exactly first to see the baseline result
• Change one element at a time: lighting, lens, setting or clothing
• Keep notes on which wording changes made the output better or worse

Use the example images as a guide, not a benchmark

Each prompt is paired with an example image from the source. Use these to judge the style you are aiming for. The README does not describe how the outputs were generated or tested, so do not treat a sample image as a guaranteed result for your own run.

• Compare your output against the example for style, not for exact match
• Expect variation between runs of the same prompt

Learn prompt wording from the examples

The strongest value for beginners is seeing how specific terms are used. Prompts in the list name camera styles such as early-2000s digital camera or 1990s-style camera with direct front flash, lighting such as harsh super-flash or soft natural fill light, and texture such as subtle grain.

Sections such as Interior Design and E-commerce show the same pattern applied to a product or room.

• Name the era or camera style explicitly
• Describe lighting and shadows, not only the subject
• Specify background and atmosphere separately from the subject