When is Nano Banana used vs GPT Image?

Nano Banana is the default for all image requests. GPT Image is used when you explicitly ask for it, or when you need readable text inside images (menus, infographics, recipe cards) — GPT Image handles text rendering better.

How much does it cost?

From $0.04 per image. Edits cost the same as generation.

Can it generate photorealistic images?

Nano Banana leans toward stylized illustration. For photorealistic results, try GPT Image or Seedream (when available).

Back to Skills

Nano Banana

Default image generation — fast text-to-image and editing with stylized illustration output

Media StackActive

Result Examples

Before

After

Coastal landscape — AI-generated Godzilla composited into a real photograph

What It Does

Nano Banana is the default image generation skill. Powered by Google Gemini, it creates stylized illustrations from text prompts and can edit existing images. Ideal for content creation, social media visuals, and creative work where speed and style matter more than photorealism.

Text-to-ImageImage EditingStylized IllustrationsFast Generation

In a Nutshell

🎨

Text to Image — generate images from text descriptions

✏️

Edit Images — modify existing images with instructions

🖼️

Stylized Output — distinctive illustration style

⚡

Fast — results in seconds

Use Cases

Social Media Visuals

Generate eye-catching graphics for posts and stories

Blog Illustrations

Create custom illustrations for articles and newsletters

Concept Exploration

Rapidly visualize ideas before committing to professional design

Quick Edits

Modify existing images — change backgrounds, add elements, adjust style

Real-world examples

Send references, get visual iterations back fast

Send references, describe tweaks. Iterations back fast with tool tips and examples.

Weekly illustrated stories for your kid

Every week my daughter gets a new illustrated story with her favorite characters.

How to Use

Step 1

Generate an image

Describe what you want. The assistant uses Nano Banana by default for all image requests.

This is the default — you don't need to say "use nano banana".

Step 2

Edit an image

Send an image and describe what you want changed. Say "edit this" or "change the background to..."

Step 3

Control style

Add style instructions to your prompt: "watercolor style", "flat illustration", "minimalist", etc.

Command Examples

You say:

Generate an illustration of a cozy coffee shop interior, warm lighting, watercolor style

Assistant responds:

MEDIA:coffee-shop-illustration.png Generated in 8s. Want me to adjust anything?

You say:

Edit this photo — replace the sky with a sunset

Assistant responds:

MEDIA:sunset-edit.png Sky replaced with sunset. The lighting on the foreground has been adjusted to match.

FAQ

Setup Requirements

Active Amplify subscription

Deposit balance for image generation