cloud-cost

Cloud Cost MCP provides multi-cloud pricing comparisons across AWS, Azure, GCP, and OCI. It supports a suite of comparison tools (compute, storage, egress, Kubernetes costs) as well as workload cost calculations and migration savings. You can interact with it using Claude Code or via the CLI by installing the cloud-cost-mcp package. Use natural language prompts to request side-by-side pricing, presets, or full workload estimates. The server surfaces real-time pricing data from public sources like instances.vantage.sh and provider APIs, enabling quick decision-making for migrations or workload planning.

Prerequisites:

• Node.js and npm installed on your system (for runtime and package management).
• Optional: Claude Code environment if you plan to integrate with Claude Code workflows.

Installation options:

1. One-command install (Claude Code users):

# One-command install
claude mcp add cloud-cost -- npx cloud-cost-mcp

2. Manual installation (CLI usage):

npm install -g cloud-cost-mcp

3. Claude Code configuration snippet (as shown in the README):

{
  "mcpServers": {
    "cloud-cost": {
      "command": "cloud-cost-mcp"
    }
  }
}

After installation, you can run the MCP via the configured command (cloud-cost-mcp) or through Claude Code integration depending on your setup.

Tips and considerations:

• This MCP pulls data from public sources (e.g., instances.vantage.sh, Azure Retail Prices API, Oracle price lists). Prices can vary by region, account type, and terms, so use the results as estimates.
• Use the real-time API checks to verify availability of pricing sources with check_api_status.
• Data freshness can be checked with get_data_freshness; ensure data is recent (watch for 30+ days age warnings).
• OCI offers notable advantages (e.g., 10TB/month free egress and free Kubernetes control plane) that can influence comparisons.
• Commands and tools are organized into: compare_compute, compare_storage, compare_egress, compare_kubernetes, find_cheapest_compute, calculate_workload_cost, quick_estimate, and more. Presets cover common scenarios like small-web-app and kubernetes-cluster.
• If you plan to deploy in a containerized environment, consider using the official npm package to ensure compatibility with your CI/CD workflows.