How to Use Nano Banana in Photoshop
Photoshop has long defined digital image manipulation, but Google’s new Nano Banana model—part of Gemini 2.5 Flash is changing how we edit. It’s a fast, detail-preserving AI now integrated directly into Photoshop’s Generative Fill tool. This tutorial walks through how to set it up, use it effectively, and balance automation with creative control.
Requirements:
- Photoshop Beta or the latest Creative Cloud version
- Adobe creative cloud subscription
- Internet connection for server-side AI processingn
Don't have a creative cloud account? Try an exernal plugin like Pablo for Photoshop instead
Step-by-step instructions
Make a Selection
Use the Lasso, Object Selection, or Mask tool to isolate areas you want to edit.
Want to edit the whole image? Press Ctrl+A (Windows) or Command+A (Mac) to select the entire image.
Access Generative Fill
With your image or selection active, click Generative Fill in the Contextual Task Bar. If you don't see it, go to Window → Contextual Task Bar to show it.
Under model options (click the Firefly icon), select Nano Banana (Gemini 2.5 Flash) instead of Firefly.
Write a prompt
Describe clearly what you want to change and what should remain the same. Use verbs like 'replace', 'change', 'move', etc.
Example: 'Replace the background with a dimly lit café interior. Keep subject lighting warm and natural.'
Use visual language—mention color tone, lighting type, or composition direction.
Generate
Click Generate to apply your prompt.
Wait for processing—usually under 30 seconds.
Iterate and Refine
If results aren’t ideal, adjust your prompt or refine your selection.
Avoid rewriting the entire prompt; tweak one variable at a time to understand model behavior.
You can re-run generations without closing your current image.
