Pinterest MCP Server
Enables image search and information retrieval from Pinterest using the Model Context Protocol. Supports searching by keywords, getting similar pins, and downloading images directly.
README
Pinterest MCP Server
A Model Context Protocol (MCP) server for Pinterest image search and information retrieval.
Features
- Search for images on Pinterest by keywords
- Retrieve detailed information about Pinterest images
- Seamless integration with Cursor IDE, Claude Desktop, and Remote MCP clients
- Dual transport modes: Stdio (Local IDEs) and HTTP/SSE (Remote Hosting / Render)
- API Key Bearer Authentication for cloud deployment
- Support for headless browser mode
- Configurable search result limits and filename templates
- Search and download images directly to local storage
Prerequisites
- Node.js (v18 or higher)
- Cursor IDE or Claude Desktop for MCP integration
Installation & Deployment
🚀 Deploying to Render (Remote MCP Server over HTTP/SSE)
You can host this Pinterest MCP server as a secure remote service on Render, Railway, or Fly.io with Bearer API Key authentication.
- Connect your repository to Render or use the included
render.yamlblueprint. - Set Environment Variables:
PORT:3000(or leave default set by Render)API_KEY:your_super_secret_key_here(Secret key required for client requests)MCP_PINTEREST_DOWNLOAD_DIR:/tmp/downloads
- Health check URL:
GET /health(Returns200 OK) - SSE connection URL:
GET /sse(RequiresAuthorization: Bearer <API_KEY>header orx-api-keyheader)
Client Request Example with Bearer Key
GET /sse HTTP/1.1
Host: pinterest-mcp.onrender.com
Authorization: Bearer your_super_secret_key_here
Unauthenticated requests receive 401 Unauthorized responses.
Local Installation via NPX (Recommended for Local Use)
npx pinterest-mcp-server
Command-line options:
# Specify download directory
npx pinterest-mcp-server --downloadDir /path/to/downloads
# Specify filename template
npx pinterest-mcp-server --filenameTemplate "pinterest_{id}"
# Specify both options
npx pinterest-mcp-server --downloadDir ./images --filenameTemplate "pinterest_{id}"
Global Installation
npm install -g pinterest-mcp-server
pinterest-mcp-server --downloadDir /path/to/downloads
Manual Installation
-
Clone repository:
git clone https://github.com/terryso/mcp-pinterest.git pinterest-mcp-server cd pinterest-mcp-server -
Install dependencies:
npm install -
Build server:
npm run build -
Start server:
npm start
Configuring as MCP Server in Cursor
- Open Cursor IDE
- Go to Settings (⚙️) > Extensions > MCP
- Click Add Server
- Enter configuration:
- Name:
Pinterest MCP - Type:
Command - Command:
node - Args:
["/path/to/mcp-pinterest/dist/pinterest-mcp-server.js"]
- Name:
Or edit your Cursor MCP config file (~/.cursor/mcp.json):
{
"pinterest": {
"command": "node",
"args": ["/path/to/mcp-pinterest/dist/pinterest-mcp-server.js"]
}
}
Environment Variables Configuration Example
{
"pinterest": {
"command": "npx",
"env": {
"MCP_PINTEREST_DOWNLOAD_DIR": "/Users/user/Desktop/Images",
"MCP_PINTEREST_FILENAME_TEMPLATE": "pin_{imageId}_{timestamp}.{fileExtension}",
"MCP_PINTEREST_PROXY_SERVER": "http://127.0.0.1:7890"
},
"args": ["pinterest-mcp-server"]
}
}
Available MCP Tools
-
pinterest_health_check: System metrics, memory usage, Chromium binary status, and write diagnostics- No required parameters.
-
pinterest_search: Search for images on Pinterest by keyword with persistent browser poolkeyword: Search term (required)limit: Number of images to return (default: 10)headless: Whether to use headless browser mode (default: true)
-
pinterest_get_similar_pins: Deep visual recommendation engine for a target Pin URL or Pin IDpin_url_or_id: Pinterest Pin URL (e.g.https://www.pinterest.com/pin/11822017768403505/) or numeric Pin ID (required)limit: Number of recommended pins to return (default: 10)headless: Whether to use headless browser mode (default: true)
-
pinterest_get_image_info: Retrieve detailed information for a Pinterest imageimage_url: Image URL (required)
-
pinterest_search_and_download: Search and download Pinterest images directly to diskkeyword: Search term (required)limit: Number of images (default: 10)headless: Whether to use headless browser mode (default: true)
Environment Variables Reference
| Variable | Description | Default |
|---|---|---|
PORT / HTTP_PORT |
HTTP Port for SSE server mode | (Stdio mode if unassigned) |
API_KEY / MCP_API_KEY |
Secret API key for HTTP/SSE authentication | (No auth if empty) |
MCP_PINTEREST_DOWNLOAD_DIR |
Root directory for downloads | ../downloads |
MCP_PINTEREST_FILENAME_TEMPLATE |
Template for downloaded file names | pinterest_{imageId}.{fileExtension} |
MCP_PINTEREST_PROXY_SERVER |
Optional proxy server (http://host:port) |
None |
PUPPETEER_EXECUTABLE_PATH |
Path to custom Chrome/Chromium binary | System default |
License
This project is licensed under the MIT License - see the LICENSE file for details.
<!-- Deployed: 2026-07-29 -->
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。
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。
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。
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。
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。
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。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
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。
