Transform images with Nano Banana


Madgwick’s new Nano Banana (aka Gemini 2.5 Flash Image) MCP brings powerful new image editing capabilities to your fingertips. Unlike traditional AI image generators, Nano Banana excels at editing and combining existing images, rather than creating new ones from scratch.

Getting started: Install your image editing toolkit

Step 1: Access the MCP Store

Click on your initials in the top right, and the MCP Store link at the top of the drop-down, or use the Box icon in the bottom left of your chatbox and click Manage Connections.

 

Look for "Nano Banana (Gemini 2.5 Flash Image)" under Image Creation.

 

Step 2: Install your core image tools

For a complete image editing workflow, install:

  • Nano Banana - for editing and combining images

  • Flux - for creating new images from text

  • GPT Image or Ideogram - alternative image generators

 

Step 2: Edit images

Drag and drop your images into the Madgwick chat, and instruct how you want them edited. The MCP image tools will create options to select from and further edit.

Understanding what Nano Banana does best

Nano Banana shines at specific tasks:

  • Compositing: Place an element from one photo into a different background

  • Style transfer: Transform a modern photo into a 1970s yearbook portrait

  • Object editing: Change colours or add elements while preserving the original image

Example workflows

Workflow 1: Creating marketing materials

Scenario: You need promotional imagery for an internal newsletter about a new project.

  1. Generate a base image with GPT Image: "Create a clean minimalist image of a steaming coffee cup next to a laptop on a modern desk"

  2. Find a UNE building image from UNE’s media library, or use Google Images' search to filter for images with licenses that allow for modification and reuse, and drag it directly to your desktop

  3. Use Nano Banana to combine: "Edit the laptop image by placing the building picture on the laptop screen instead of the text"

Workflow 2: Generate podcast cover image

Scenario: Converting research materials into digestible formats for students.

  1. Use the Academic Search MCP to gather papers on your topic

  2. Switch to your Podcast Assistant in the same chat

  3. Generate a podcast summary of the research

  4. Create cover art using Nano Banana to combine relevant images

Workflow 3: Multi-step project imagery

Scenario: Building visual content for extended projects.

  1. Create a dedicated project thread in Madgwick

  2. Dump all relevant images, emails, and context into the thread

  3. Use Flux to generate new backgrounds

  4. Use Nano Banana to composite existing elements into those backgrounds

  5. Track image tasks with the Task Planner MCP

Pro tips

  • Use multiple tabs: Run different image tasks simultaneously. Have Flux generating in one tab while Nano Banana edits in another.

  • Maintain project threads: Keep all images for a project in one chat thread. With Gemini 2.5 Pro's million-token context, you can work on the same project for weeks.

  • Try a one-shot approach: Where the image edit requires a generic addition rather than a specific image, try asking Nano Banana to add it outright. If high resolution images aren’t required, it's adept at blending its own elements - taking into account important specifics of the base image like lighting and perspective.


Find out more by viewing the Madgwick Community of Practice presentation (starts at 46 minutes) or checking our MCP guide.

 
Previous
Previous

Build AI learning experiences with Cogniti Assist

Next
Next

Madgwick update: MCPs connect AI to tools