How to Add Nano Banana Image Tools to Claude Desktop, VS Code, and Cursor with MCP

If you already use an AI coding assistant every day, the slow part of image work is often not the model itself. It is the switching: open a separate product, paste a prompt, upload references, wait, copy the result, and then return to the conversation where the idea started. The Nano Banana MCP setup solves that by exposing image generation and editing as tools inside MCP-capable clients such as Claude Desktop, VS Code, and Cursor.

What you can do

The Nano Banana MCP Server is designed around practical image workflows rather than a single text-to-image button. According to the integration guide, it supports:

• Image generation: create high-quality images from text prompts.
• Image editing: modify an existing image or combine multiple images.
• Virtual try-on: dress clothing onto photos of people.
• Product placement: place a product into a real scene.
• Multi-model selection: use nano-banana, nano-banana-2, or nano-banana-pro.
• Task querying: check generation progress and retrieve results.

The important builder detail is that these capabilities become callable tools in your AI client. Instead of treating image generation as a disconnected web form, you can describe the outcome in the same chat where you are writing UI copy, preparing a product page, or iterating on visual assets.

How it works

MCP, or Model Context Protocol, lets an AI client call external tools through a standardized interface. In this setup, the client starts a local command named mcp-nanobanana-pro. That command exposes Nano Banana tools to the client, and it receives credentials through the ACEDATACLOUD_API_TOKEN environment variable.

The shape is simple:

1. Install the MCP server package.
2. Add a server entry to your AI client configuration.
3. Pass your Ace Data Cloud API token through env.
4. Restart the client so it discovers the new tools.
5. Ask for image generation, editing, try-on, product placement, or task status in natural language.

The guide documents four tool names exposed by the server: nanobanana_generate_image, nanobanana_edit_image, nanobanana_get_task, and nanobanana_get_tasks_batch. For many teams, those four operations are enough to cover the loop from prompt to asynchronous result checking.

Install the Nano Banana MCP server

The recommended installation path is pip:

pip install mcp-nanobanana-pro

If you prefer working directly from source, the guide also documents a Git-based installation:

git clone https://github.com/AceDataCloud/NanoBananaMCP.git
cd NanoBananaMCP
pip install -e .

After installation, the command you configure in your MCP client is mcp-nanobanana-pro. Keep that name exact; it is what the client will execute when loading the server.

Configure Claude Desktop

For Claude Desktop, edit the client configuration file. The documented locations are:

• macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
• Windows: %APPDATA%\Claude\claude_desktop_config.json

Add a server entry named nanobanana under mcpServers:

{
  "mcpServers": {
    "nanobanana": {
      "command": "mcp-nanobanana-pro",
      "env": {
        "ACEDATACLOUD_API_TOKEN": "Your API Token"
      }
    }
  }
}

If you use uvx and do not want to install the package in advance, the guide provides this variant:

{
  "mcpServers": {
    "nanobanana": {
      "command": "uvx",
      "args": ["mcp-nanobanana-pro"],
      "env": {
        "ACEDATACLOUD_API_TOKEN": "Your API Token"
      }
    }
  }
}

Save the file and restart Claude Desktop. After restart, the conversation can access the Nano Banana tools.

Configure VS Code or Cursor

For VS Code or Cursor, create .vscode/mcp.json in the project root. The structure is slightly different from Claude Desktop: it uses servers instead of mcpServers.

{
  "servers": {
    "nanobanana": {
      "command": "mcp-nanobanana-pro",
      "env": {
        "ACEDATACLOUD_API_TOKEN": "Your API Token"
      }
    }
  }
}

The uvx version is also supported:

{
  "servers": {
    "nanobanana": {
      "command": "uvx",
      "args": ["mcp-nanobanana-pro"],
      "env": {
        "ACEDATACLOUD_API_TOKEN": "Your API Token"
      }
    }
  }
}

This project-level configuration is useful when image work belongs to a specific repository: for example, a landing page, design system, marketplace listing, or internal content pipeline.

Use it like a builder, not a demo

Once the server is configured, you do not need to memorize every tool name. The guide’s examples are natural-language tasks such as generating a watercolor landscape, putting clothing onto a person, placing a product in a café scene, or generating a high-quality portrait with the nano-banana-pro model.

A practical workflow might look like this:

• Ask the assistant to generate a product-context image for a landing page.
• Provide a reference image and ask it to edit or combine assets.
• Use task querying when the result is asynchronous.
• Iterate from the same conversation where you are also writing copy or code.

The setup is intentionally small: one package, one command, one token environment variable, and a client config file. That is why MCP is a good fit here. It lets image generation live next to the rest of your development workflow instead of becoming another disconnected tab.

Read the source integration guide here: Nano Banana MCP Integration Guide.