jae-jae/fetcher-mcp
Free使用 Playwright 无头浏览器获取网页内容，支持 JS 渲染和智能提取 (Markdown/HTML)。
FreeFree tier
Inputs: url
About jae-jae/fetcher-mcp
Fetcher MCP is an open-source MCP server that leverages Playwright's headless browser to fetch and extract web page content. It executes JavaScript to handle dynamic web applications, intelligently extracts main content using the Readability algorithm (removing ads and navigation), and supports both HTML and Markdown output. Key features include parallel URL fetching for batch operations, automatic blocking of unnecessary resources (images, stylesheets, fonts) for bandwidth efficiency, robust error handling, and configurable timeouts and output formatting. It can be run directly via npx with automatic browser installation, or deployed using Docker. The server supports Streamable HTTP and SSE endpoints, and includes a debug mode that shows the browser window.
Key Features
JavaScript support: Uses Playwright to execute JavaScript, handling dynamic web content and modern web applications.
Intelligent content extraction: Built-in Readability algorithm extracts main content, removing ads, navigation, and non-essential elements.
Flexible output format: Supports both HTML and Markdown output formats for easy integration.
Parallel processing: fetch_urls tool enables concurrent fetching of multiple URLs for batch operations.
Resource optimization: Automatically blocks unnecessary resources (images, stylesheets, fonts, media) to reduce bandwidth usage.
Robust error handling: Comprehensive error handling and logging ensures reliable operation.
Configurable parameters: Fine-grained control over timeouts, content extraction, and output formatting.
Multiple transport modes: Supports Streamable HTTP and SSE endpoints for flexible client connection.
Debug mode: Option to show browser window for debugging with --debug flag.
Docker deployment: Easy deployment via Docker with docker-compose support.
Pros & Cons
Pros
- Handles JavaScript-heavy sites that traditional scrapers cannot.
- Intelligent content extraction removes clutter for cleaner data.
- Supports parallel fetching for efficiency.
- Resource blocking reduces bandwidth and speeds up requests.
- Configurable timeouts and output formatting for flexibility.
- Free and open-source with easy setup via npx or Docker.
- Actively maintained with good documentation and community support.
Cons
- Requires browser binary download (Playwright) on first run, which can be slow.
- Headless browser usage may be overkill for simple static pages.
- Primarily designed for MCP protocol clients; less useful as a standalone tool without an MCP host.
- No built-in support for authentication or session handling beyond cookies? Not explicitly mentioned.
- Resource blocking may break some sites that rely on CSS or images for functionality.
Best For
Scraping dynamic JavaScript-rendered web pages for data extraction.Feeding clean, extracted web content into LLMs or downstream applications.Batch processing multiple URLs for content aggregation or monitoring.Integrating web fetching capabilities into AI agent workflows (via MCP protocol).Automating website content analysis with structured output (Markdown/HTML).
FAQ
How do I install Fetcher MCP?
You can run it directly using npx: npx -y fetcher-mcp. On first run, you need to install the required browser by executing: npx playwright install chromium. Alternatively, you can deploy via Docker using the provided Docker image.
What output formats does Fetcher MCP support?
It supports both HTML and Markdown output formats, making it easy to integrate with various downstream applications.
Can Fetcher MCP fetch multiple URLs at once?
Yes, the fetch_urls tool enables concurrent fetching of multiple URLs, significantly improving efficiency for batch operations.
What is the difference between the HTTP and SSE endpoints?
Fetcher MCP provides a Streamable HTTP endpoint (/mcp) using the modern MCP protocol, and a legacy SSE endpoint (/sse) for backwards compatibility. Clients can choose either method.
Is Fetcher MCP free to use?
Yes, Fetcher MCP is open-source and free to use. It is licensed under an open-source license (see the repository for details).