Dumpling AI MCP
DumplingAI MCP Server integrates with Dumpling AI for data scraping, AI tasks, and code execution. It offers web scraping, document conversion, knowledge management, and developer tools.
How to Install and Use Dumpling AI MCP
Dumpling AI MCP is a powerful tool that helps you work with lots of data and AI tasks like web scraping, document processing, and running AI agents. Here’s a simple guide on how to get it running and use it with some basic examples.
Installing Dumpling AI MCP
First, you need to install the tool. There are a few ways to do this, but here is the easiest:
-
Using npx to run without full install
You can run Dumpling AI MCP quickly using this command:
env DUMPLING_API_KEY=your_api_key npx -y mcp-server-dumplingaiReplace
your_api_keywith your actual Dumpling AI API key. This command will start the server for you right away. -
Manual global install
If you want to install it globally on your computer so it's always available, run this:
npm install -g mcp-server-dumplingai -
Installing via Smithery
If you are using Claude Desktop, you can install it automatically by running:
npx -y @smithery/cli install @Dumpling-AI/mcp-server-dumplingai --client claude
Setting Up Environment Variables
Before running Dumpling AI MCP, you need to set your API key. The main environment variable you must set is:
DUMPLING_API_KEY: Your Dumpling AI API key (it’s required).
This key gives Dumpling AI MCP the permission to connect to the API and work properly.
Running Dumpling AI MCP with Cursor
If you use the Cursor app, you can add Dumpling AI MCP as a tool there:
- Open Cursor Settings.
- Go to Features > MCP Servers.
- Click Add New MCP Server.
- Enter this JSON configuration:
Replace{ "mcpServers": { "dumplingai": { "command": "npx", "args": ["-y", "mcp-server-dumplingai"], "env": { "DUMPLING_API_KEY": "<your-api-key>" } } } }<your-api-key>with your actual API key.
Note: If you have problems on Windows, try this command instead:
cmd /c "set DUMPLING_API_KEY=your-api-key && npx -y mcp-server-dumplingai"
Example: Fetch YouTube Video Transcript
Once Dumpling AI MCP is running, you can use its tools. For example, to get a transcript from a YouTube video, send this JSON:
{
"name": "get-youtube-transcript",
"arguments": {
"videoUrl": "https://www.youtube.com/watch?v=dQw4w9WgXcQ",
"includeTimestamps": true,
"timestampsToCombine": 3,
"preferredLanguage": "en"
}
}
This command extracts the transcript with timestamps combined in groups of 3 and in English.
Example: Search Google for Machine Learning Basics
To perform a web search and scrape results in markdown format, use:
{
"name": "search",
"arguments": {
"query": "machine learning basics",
"country": "us",
"language": "en",
"dateRange": "pastMonth",
"scrapeResults": true,
"numResultsToScrape": 3,
"scrapeOptions": {
"format": "markdown",
"cleaned": true
}
}
}
This fetches recent search results and scrapes the content cleanly in markdown form.
By following these simple steps, you can install Dumpling AI MCP and start using its many tools for web scraping, AI completions, and data extraction. Just be sure you have your Dumpling API key ready and use the commands exactly as shown.