AWS Nova Canvas

MCP server for generating and editing images using AWS Bedrock’s Nova Canvas model

python

4

GitHub Stars

python

Language

6 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "yunwoong7-aws-nova-canvas-mcp": {
      "command": "uvx",
      "args": [
        "aws-nova-canvas-mcp"
      ],
      "env": {
        "AWS_PROFILE": "YOUR_AWS_PROFILE"
      }
    }
  }
}

You can run the AWS Nova Canvas MCP Server to generate and edit images using the Nova Canvas model for Bedrock. This server exposes a set of image generation features and keeps produced images organized on your computer. Follow these steps to install, start, and use the MCP server with your MCP client.

How to use

Use a compatible MCP client to connect to the AWS Nova Canvas MCP Server. Start the server using the runtime command and then issue image-generation requests through the client’s MCP interface. You can perform text-to-image generation, inpainting, outpainting, image variation, conditioning, color-guided generation, and background removal, with thumbnails available for quick preview.

Starting the server allows you to manage all image generation tasks from your MCP client, and generated images get saved to your local machine by default. Use the client’s controls to select prompts, reference images, colors, and other parameters to achieve your desired output.

How to install

Prerequisites you need before starting:

Install the MCP runtime tooling used to run local MCP servers.

Configure your AWS credentials so the server can access Bedrock resources. You can set your AWS profile using the AWS_PROFILE environment variable.

Install and prepare to run the server using the exact command shown for this MCP configuration.

uvx aws-nova-canvas-mcp

Additional notes and configuration

Environment variable you must provide: AWS_PROFILE. Use YOUR_AWS_PROFILE as a placeholder until you replace it with your actual AWS profile name.

Default save location for generated or edited images varies by operating system. Use these paths to locate your images if you want to copy or back them up immediately:

macOS / Linux: ~/Desktop/aws-nova-canvas Windows: C:\Users\YourUsername\Desktop\aws-nova-canvas

If you configure a different save path in your client or environment, ensure the path exists or the server can create it automatically.

Important: MCP settings and the integration described here are intended for Claude desktop environments that support the MCP flow. Ensure your client and runtime versions are compatible with this MCP configuration.

Troubleshooting tips

Verify AWS_PROFILE is set and points to a valid AWS configuration with permissions for Bedrock access.
Confirm uvx is installed and accessible from your shell path.
Check that the MCP server process starts without errors and remains running while the client is connected.
Ensure the save directory exists or that the server has permission to create it if needed.

Available tools

Text to Image

Generate images from text prompts using the Nova Canvas model.

Image Inpainting

Fill or modify parts of an image by guiding the model with a prompt or mask.

Image Outpainting

Extend an image beyond its original borders to create larger compositions.

Image Variation

Create variations of an existing image by applying stylistic or content changes.

Image Conditioning

Influence generation with conditioning inputs such as reference images or attributes.

Color Guided Generation

Guide color palettes and color relationships in generated images.

Background Removal

Remove backgrounds from images to isolate subjects.

Show Image Thumbnails

Preview generated or edited images as thumbnails for quick review.