Skip to main content
Scrapling’s MCP server provides six powerful tools for web scraping operations. Each tool is optimized for different use cases and protection levels.

​
Available Tools

​
get

Make stealth HTTP GET requests to fetch web pages. Best for: Low to mid protection levels, simple HTTP requests Parameters:
​
url
string
required
The URL to request
​
impersonate
string
default:"chrome"
Browser to impersonate (chrome, firefox, safari, etc.)
​
extraction_type
string
default:"markdown"
Output format: markdown, html, or text
​
css_selector
string
CSS selector to extract specific content
​
main_content_only
boolean
default:"true"
Extract only content within <body> tag
​
headers
object
Custom HTTP headers
​
cookies
object
Cookies to include in request
​
proxy
string
Proxy URL (format: “http://user:pass@host:port”)
​
timeout
number
default:"30"
Request timeout in seconds
​
stealthy_headers
boolean
default:"true"
Use real browser headers
Example usage: 
{
  "url": "https://example.com",
  "extraction_type": "markdown",
  "css_selector": "article.main",
  "impersonate": "chrome"
}

​
bulk_get

Fetch multiple URLs concurrently with HTTP GET requests. Best for: Scraping multiple pages efficiently Parameters: Same as get, but accepts urls (array) instead of url (string).
​
urls
array[string]
required
List of URLs to fetch concurrently
Example usage: 
{
  "urls": [
    "https://example.com/page1",
    "https://example.com/page2",
    "https://example.com/page3"
  ],
  "extraction_type": "markdown",
  "impersonate": "firefox"
}

​
fetch

Use Playwright browser automation for JavaScript-heavy sites. Best for: Single-page applications, sites requiring JavaScript execution Parameters:
​
url
string
required
The URL to fetch
​
extraction_type
string
default:"markdown"
Output format: markdown, html, or text
​
headless
boolean
default:"true"
Run browser in headless mode
​
disable_resources
boolean
default:"false"
Block images, fonts, media for speed boost
​
network_idle
boolean
default:"false"
Wait for no network activity for 500ms
​
timeout
number
default:"30000"
Timeout in milliseconds
​
wait
number
default:"0"
Additional wait time in milliseconds
​
wait_selector
string
CSS selector to wait for before proceeding
​
wait_selector_state
string
default:"attached"
State to wait for: attached, detached, visible, hidden
​
real_chrome
boolean
default:"false"
Use real Chrome installation instead of Chromium
​
google_search
boolean
default:"true"
Set referer as Google search of domain
Example usage: 
{
  "url": "https://spa-website.com",
  "extraction_type": "markdown",
  "wait_selector": "div.content-loaded",
  "network_idle": true,
  "disable_resources": true
}

​
bulk_fetch

Fetch multiple URLs with browser automation concurrently. Best for: Scraping multiple JavaScript-heavy pages Parameters: Same as fetch, but accepts urls (array) instead of url (string). Example usage: 
{
  "urls": [
    "https://app1.example.com",
    "https://app2.example.com"
  ],
  "headless": true,
  "network_idle": true
}

​
stealthy_fetch

Advanced stealth browser automation with Cloudflare bypass. Best for: High protection sites, Cloudflare-protected pages Parameters: All fetch parameters, plus:
​
solve_cloudflare
boolean
default:"false"
Automatically solve Cloudflare challenges
​
block_webrtc
boolean
default:"false"
Block WebRTC to prevent IP leaks
​
allow_webgl
boolean
default:"true"
Allow WebGL (recommended for stealth)
​
hide_canvas
boolean
default:"false"
Add noise to canvas fingerprinting
​
additional_args
object
Additional Playwright context settings
Example usage: 
{
  "url": "https://protected-site.com",
  "extraction_type": "markdown",
  "solve_cloudflare": true,
  "block_webrtc": true,
  "hide_canvas": true,
  "wait": 2000
}

​
bulk_stealthy_fetch

Fetch multiple protected URLs with advanced stealth. Best for: Scraping multiple Cloudflare-protected sites Parameters: Same as stealthy_fetch, but accepts urls (array) instead of url (string). Example usage: 
{
  "urls": [
    "https://protected1.com",
    "https://protected2.com"
  ],
  "solve_cloudflare": true,
  "network_idle": true
}

​
Response Format

All tools return a structured response: 
{
  "status": 200,
  "content": ["Extracted content in requested format"],
  "url": "https://example.com"
}
For bulk operations, an array of responses is returned: 
[
  {
    "status": 200,
    "content": ["Content from URL 1"],
    "url": "https://example.com/page1"
  },
  {
    "status": 200,
    "content": ["Content from URL 2"],
    "url": "https://example.com/page2"
  }
]

​
Extraction Types

Converts HTML to clean Markdown format:
{"extraction_type": "markdown"}
Best for: Readable text, content processing, AI consumption

​
CSS Selectors

All tools support CSS selectors for targeted extraction: 
# Extract all articles
{"css_selector": "article.post"}

# Extract main content
{"css_selector": "main#content"}

# Extract specific elements
{"css_selector": "div.product-info"}
When css_selector matches multiple elements, all matches are returned in the content array.

​
Authentication

​
HTTP Basic Auth

{
  "url": "https://example.com",
  "auth": {
    "username": "user",
    "password": "pass"
  }
}

​
Proxy Authentication

{
  "url": "https://example.com",
  "proxy": "http://proxy.example.com:8080",
  "proxy_auth": {
    "username": "proxy_user",
    "password": "proxy_pass"
  }
}

​
Common Patterns

{
  "url": "https://example.com",
  "extraction_type": "markdown"
}

{
  "url": "https://news.example.com/article",
  "css_selector": "article.content",
  "extraction_type": "markdown",
  "main_content_only": true
}

{
  "url": "https://spa.example.com",
  "wait_selector": "div.loaded",
  "network_idle": true,
  "extraction_type": "html"
}

{
  "url": "https://protected.example.com",
  "solve_cloudflare": true,
  "wait": 2000,
  "extraction_type": "markdown"
}

{
  "urls": [
    "https://site1.com",
    "https://site2.com",
    "https://site3.com"
  ],
  "impersonate": "chrome,firefox,safari",
  "stealthy_headers": true,
  "extraction_type": "markdown"
}

​
Tool Selection Guide

1

Simple HTTP sites

Use get or bulk_get for basic HTML pages without JavaScript
2

JavaScript-heavy sites

Use fetch or bulk_fetch for SPAs and dynamic content
3

Protected sites

Use stealthy_fetch or bulk_stealthy_fetch for Cloudflare and WAF-protected sites
4

Multiple URLs

Use bulk variants (bulk_get, bulk_fetch, bulk_stealthy_fetch) for concurrent operations

MCP Server

Learn about the MCP server

Setup Guide

Configure MCP server for AI clients

Build docs developers (and LLMs) love

Get started for freeTalk to us