aws-calculator-mcp
Enables creating AWS cost estimates through natural language, generating shareable calculator.aws links, and retrieving actual calculated costs via headless Chrome, without requiring AWS credentials.
README
aws-calculator-mcp
MCP server for AWS Pricing Calculator. Create cost estimates through natural language, get shareable links, retrieve actual AWS-calculated costs.
- 436 services, 72 regions
- Real pricing via headless Chrome (not approximations)
- Savings Plans, Reserved Instances, all pricing models
- No AWS credentials required
Install
Add to your MCP client config:
{
"mcpServers": {
"aws-calculator": {
"command": "npx",
"args": ["-y", "aws-calculator-mcp@latest"]
}
}
}
| Client | Config file |
|---|---|
| Kiro | ~/.kiro/settings/mcp.json |
| Claude Desktop (macOS) | ~/Library/Application Support/Claude/claude_desktop_config.json |
| Claude Desktop (Windows) | %APPDATA%\Claude\claude_desktop_config.json |
| Cursor | .cursor/mcp.json |
| VS Code | .vscode/mcp.json |
Requires Node.js ≥ 18. Chrome optional (only for cost retrieval).
What it does
You ask your AI assistant to estimate AWS costs. The MCP handles the rest:
- Searches the AWS service catalog
- Looks up valid configuration fields
- Builds the estimate incrementally
- Exports to calculator.aws (shareable link)
- Opens the link in headless Chrome to get actual calculated costs
The pricing is identical to calculator.aws because it is calculator.aws doing the math.
Tools
| Tool | Purpose |
|---|---|
search_services |
Find services by keyword |
get_service_fields |
Get valid config fields for a service |
create_estimate |
Start a new estimate |
add_service |
Add configured services |
update_service |
Modify a service in-place |
export_estimate |
Get a shareable calculator.aws URL |
refresh_estimate |
Get actual costs via headless Chrome |
generate_report |
Per-service cost breakdown (CSV/Markdown) |
import_estimate |
Download an existing estimate |
Example
"Estimate a production stack in us-west-2: 2x c5.xlarge EC2 with Savings Plan, Lambda at 20M requests, S3 500GB, CloudFront 1TB"
→ Returns a calculator.aws link + $333.92/mo breakdown by service.
How it works
Your prompt → Agent configures services → MCP packages payload → calculator.aws saves it
→ Chrome opens it → scrapes real costs
No local price calculations. No hardcoded rates. The AWS calculator frontend computes everything.
Requirements
- Node.js ≥ 18
- Chrome (optional) — for
refresh_estimateandgenerate_report. Install withnpx puppeteer browsers install chromeor use system Chrome.
License
MIT
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。
mcp-server-qdrant
这个仓库展示了如何为向量搜索引擎 Qdrant 创建一个 MCP (Managed Control Plane) 服务器的示例。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。