AWS EC2 Pricing MCP Server – exposes pre-parsed EC2 price catalogue through the MCP protocol so LLMs/scripts can query by RAM, CPU, term, OS, etc.
https://github.com/trilogy-group/aws-pricing-mcpYou know the drill. You need to find the most cost-effective EC2 instance for your workload, but AWS pricing is a maze. You're bouncing between the pricing calculator, digging through instance family documentation, and trying to mentally calculate reserved instance discounts across regions.
The AWS EC2 Pricing MCP Server changes that completely. Ask it natural questions like "What's the cheapest instance with 32GB RAM?" and get instant, accurate answers.
This MCP server exposes AWS's entire EC2 pricing catalog through a pre-parsed interface that understands what you actually want to know:
No more context switching between AWS documentation, pricing pages, and calculators. Your LLM assistant becomes a pricing expert with real-time data.
Here's what sets this apart: the entire EC2 pricing catalog is pre-processed and cached locally. When you ask about pricing, you're not waiting for AWS API calls or parsing complex JSON responses. You get sub-second answers from a structured, queryable dataset.
The server maintains fresh pricing data automatically, pulling updates every 24 hours while serving queries from the local cache. Your cost analysis workflows stay fast and reliable.
This server shines in scenarios where accurate, fast pricing data drives decisions:
Cost Optimization Audits: "Show me all instances in our current stack that could be replaced with cheaper alternatives having the same CPU/RAM specs."
Multi-Region Planning: Compare pricing across regions for your specific requirements without manually checking each region's pricing page.
Reserved Instance Analysis: Get exact savings calculations for different commitment terms and payment options across instance families.
Rightsizing Recommendations: Find instances that match your actual resource needs instead of over-provisioning based on incomplete pricing knowledge.
Docker Option (Recommended):
{
"mcpServers": {
"AWS EC2 Pricing MCP": {
"command": "docker",
"args": ["run", "--rm", "-i", "--network", "none", "ai1st/aws-pricing-mcp"]
}
}
}
The --network none
flag ensures complete network isolation - the container can't make external calls, giving you security confidence when processing sensitive infrastructure data.
Direct Python: Download the pricing data once, then run queries locally without any network dependencies.
Works immediately with Claude Desktop, Continue, or any MCP-compatible tool. Your existing LLM workflows get instant access to comprehensive AWS pricing intelligence without changing how you work.
The server handles the complexity of AWS's pricing structure - different regions, instance families, operating systems, and licensing options - so your queries stay simple and conversational.
This isn't just another API wrapper. It's pricing data designed for the way developers actually think about infrastructure costs.