Aws Pricing Mcp

AWS EC2 Pricing MCP Server

The AWS EC2 Pricing MCP Server lets any LLM or automation script query real-time EC2 pricing with one call. Powered by a pre-parsed AWS pricing catalogue, it answers questions such as

What is the cheapest EC2 instance with 32GB RAM?

Which AMD instances have more than 3.5 Ghz CPUs?

What is the 3-yr All Upfront discount on r6g family in eu-west-1?

What is the cheapest instance to run Windows with SQL Server Enterprise?

Using Docker

Using a Docker hub image

Use this mcp_config.json for Docker hub image:

{
  "mcpServers": {
    "AWS EC2 Pricing MCP": {
      "command": "docker",
      "args": ["run", "--rm", "-i", "-q", "--network", "none", "ai1st/aws-pricing-mcp"]
    }
  }
}

• The --rm flag removes the container when it exits.
• The -i flag enables interactive mode for stdio communication.
• The -q flag suppresses the docker messages about downloading the image.
• The --network none totally disconnects the container from the network to guarantee no data exfiltration.

Using a local image

Build the image:

docker build -t aws-pricing-mcp . --build-arg BUILD_DATE=$(date +%Y-%m-%d)

This will download the pricing data and build the image. The BUILD_DATE parameter ensures the fresh pricing data is downloaded during build.

Sample mcp_config.json for a locally built image:

{
  "mcpServers": { "AWS EC2 Pricing MCP": {
      "command": "docker",
      "args": ["run", "--rm", "-i", "--network", "none", "aws-pricing-mcp"]
    }
  }
}

• The --rm flag removes the container when it exits.
• The -i flag enables interactive mode for stdio communication.
• The --network none totally disconnects the container from the network to guarantee no data exfiltration.

Using Python directly

You’ll need to download the pricing data first:

curl https://cloudfix-public-aws-pricing.s3.us-east-1.amazonaws.com/pricing/ec2_pricing.json.gz | gunzip > ec2_pricing.json

It should be in the same directory as server.py.

Sample mcp_config.json for local Python:

{
  "mcpServers": { "AWS EC2 Pricing MCP": {
      "command": "python",
      "args": [
        "/path/to/server.py"
      ]
    }
  }
}

Building Instructions

For instructions on building and publishing the Docker image, see BUILD.md{:target=“_blank”}.

Pricing Data JSON Format

See PRICING.md{:target=“_blank”}.