AWS Pricing MCP

Overview

The AWS Pricing MCP server offers a streamlined, protocol‑compliant gateway for retrieving up‑to‑date Amazon EC2 pricing information. By exposing this data through the Model Context Protocol, AI assistants can query and filter cloud costs in real time without hard‑coding pricing tables or maintaining local caches. The server supports both a conventional HTTP endpoint and a fully serverless Lambda deployment, giving teams the flexibility to choose between persistent service or cost‑effective, on‑demand execution.

At its core, the MCP server pulls pricing snapshots from Amazon S3 and presents them in a JSON‑RPC 2.0 compliant format that matches the MCP specification. Clients can request instance types, regions, tenancy options, and platform details, then filter results by vCPU count, memory size, GPU presence, or other attributes. The server automatically translates these filters into the appropriate AWS pricing dimensions, returning a ranked list of the cheapest options for any given workload profile. This eliminates manual look‑ups in the AWS Pricing Calculator and ensures that AI assistants always work with the latest market rates.

Key capabilities include:

• Multi‑model pricing: On Demand, Reserved Instances, and CloudFix RightSpend are all available in a single query, allowing assistants to recommend the most cost‑effective purchasing strategy.
• Fine‑grained filtering: Region, platform (Linux/Windows), tenancy (shared or dedicated), and compute specifications can be combined to narrow results.
• Dynamic data refresh: Pricing files are sourced directly from S3, so the server reflects changes as soon as new snapshots are published.
• Protocol compliance: The JSON‑RPC 2.0 interface adheres to the MCP spec, enabling seamless integration with Claude or other AI assistants that understand MCP.

Typical use cases span a range of developer and operations scenarios:

• Cost‑aware code generation: An AI assistant can suggest the most economical instance type for a new microservice, taking into account region and workload requirements.
• Automated deployment scripts: Infrastructure-as‑Code tools can query the MCP server during CI/CD pipelines to pick the cheapest instance configuration before provisioning.
• Financial forecasting: Teams can embed pricing queries into budgeting tools, letting AI assistants predict monthly spend based on projected usage.

The server’s dual deployment model is a standout feature. The Lambda function, accessible via an HTTPS Function URL, removes the need to maintain a persistent server and automatically scales with traffic. The traditional Python server remains an option for environments that require persistent connections or custom middleware. By combining protocol‑level standardization with cloud‑native deployment options, the AWS Pricing MCP server empowers developers to weave accurate, real‑time pricing intelligence into AI‑driven workflows effortlessly.