home / mcp / nanobanana mcp server

NanoBanana MCP Server

Provides image generation and editing via NanoBanana through AceDataCloud API for MCP clients.

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "acedatacloud-mcpnanobanana": {
      "url": "https://api.acedata.cloud/mcp/nanobanana",
      "headers": {
        "ACEDATACLOUD_API_TOKEN": "YOUR_API_TOKEN",
        "ACEDATACLOUD_API_BASE_URL": "https://api.acedata.cloud"
      }
    }
  }
}

You run an MCP server that talks to the NanoBanana AI image model through the AceDataCloud API. This lets you generate and edit images, try on clothing virtually, place products in scenes, and track tasks from any MCP-compatible client like Claude or VS Code.

How to use

Connect to the MCP server from your client by starting the server locally and pointing your client to the MCP instance. You will authenticate with an API token and then call image generation or editing endpoints exposed by the server. Use the available tools to generate images from prompts, edit existing images, or compose scenes like product placements. Tasks such as querying the status of a generation are also available.

How to install

Prerequisites: you need Python installed on your system. You also need access to the AceDataCloud platform to obtain your API token.

1) Clone the project and move into it.

2) Install the package in editable mode.

Concrete installation steps

# Clone the repository
git clone https://github.com/AceDataCloud/MCPNanoBanana.git
cd MCPNanoBanana

# Install with pip
pip install -e .

# Or with uvx (recommended for development)
uvx pip install -e .

Configure the API token

# Copy example environment file
cp .env.example .env

# Edit with your API token
echo "ACEDATACLOUD_API_TOKEN=your_token_here" > .env

Run the server

# Run the server
mcp-nanobanana-pro

# Or run directly with Python
python main.py

Optional Claude Desktop integration

You can configure Claude Desktop to connect to this MCP server. Use the following example configuration to define the server and provide your API token.

{
  "mcpServers": {
    "nanobanana": {
      "command": "mcp-nanobanana-pro",
      "env": {
        "ACEDATACLOUD_API_TOKEN": "your_api_token_here"
      }
    }
  }
}

Alternative runtime with uv

If you prefer uv, you can run the MCP server through uv by providing the directory and command similar to the example below.

{
  "mcpServers": {
    "nanobanana": {
      "command": "uv",
      "args": ["run", "--directory", "/path/to/MCPNanoBanana", "mcp-nanobanana-pro"],
      "env": {
        "ACEDATACLOUD_API_TOKEN": "your_api_token_here"
      }
    }
  }
}

Available tools

nanobanana_generate_image

Generate an image from a text prompt using the NanoBanana model.

nanobanana_edit_image

Edit or combine images with AI using the NanoBanana toolset.

nanobanana_get_task

Query a single task status for an in-progress or completed image operation.

nanobanana_get_tasks_batch

Query multiple task statuses in a single call.