Apify MCP

Overview

Apify MCP is a Model Context Protocol server that provides access to Apify, a web scraping and automation platform with over 2,500 pre-built scrapers (actors). Apify enables AI agents to extract data from websites, automate browser interactions, and run web scraping jobs at scale.

With pre-built scrapers for popular platforms like Instagram, TikTok, Google Maps, Amazon, and thousands more, Apify MCP lets AI agents leverage battle-tested scrapers without writing custom scraping code.

Features

• 2,500+ Pre-Built Scrapers: Ready-to-use actors for popular websites
• Browser Automation: Headless Chrome/Firefox scraping
• Proxy Rotation: Built-in proxy management for anti-bot bypass
• Queue Management: Handle pagination and link following
• Data Export: JSON, CSV, Excel, and more
• Schedule Jobs: Run scrapers on a schedule
• API Access: Programmatic control of all Apify features
• Scalable Infrastructure: Cloud-based scraping at scale

Installation

npx -y @modelcontextprotocol/server-apify

Configuration

Add to your claude_desktop_config.json:

{
  "mcpServers": {
    "apify": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-apify"],
      "env": {
        "APIFY_API_TOKEN": "YOUR_APIFY_API_TOKEN"
      }
    }
  }
}

Getting an Apify API Token

1. Go to Apify Console
2. Sign up or log in
3. Go to Account → API tokens
4. Create a new API token
5. Add the token to your MCP configuration

Available Tools

Tool	Description
run_actor	Run a pre-built scraper (actor)
get_dataset	Retrieve scraped data from a run
get_request_queue	Get request queue status
list_actors	List available pre-built scrapers
search_actors	Search for actors by keyword
get_actor_details	Get actor documentation and inputs

Usage Examples

Run a Pre-Built Scraper

# Run Google Maps scraper
run_result = run_actor(
    actor_id="apify/google-maps-scraper",
    input={
        "searchQueries": ["restaurants in New York", "cafes in San Francisco"],
        "maxResults": 50,
    }
)
print(f"Run ID: {run_result.run_id}")
print(f"Status: {run_result.status}")

Get Scraped Data

# Get results from a run
dataset = get_dataset(run_id=run_result.run_id)
for item in dataset.items:
    print(f"{item.name}: {item.address}")
    print(f"  Rating: {item.rating}")
    print(f"  Phone: {item.phone}")

Search for Actors

# Search for Instagram scrapers
actors = search_actors(query="instagram")
for actor in actors:
    print(f"{actor.name}: {actor.description}")
    print(f"  Uses: {actor.usage_count} runs")

Instagram Scraper Example

# Scrape Instagram profile
run_result = run_actor(
    actor_id="apify/instagram-scraper",
    input={
        "startUrls": [{"url": "https://www.instagram.com/nasa/"}],
        "profileResults": True,
        "postsLimit": 50,
    }
)

# Get results
posts = get_dataset(run_id=run_result.run_id)
for post in posts.items:
    print(f"Post: {post.shortUrl}")
    print(f"  Likes: {post.likes}")
    print(f"  Caption: {post.caption[:100]}...")

E-Commerce Product Scraper

# Scrape Amazon products
run_result = run_actor(
    actor_id="apify/amazon-scraper",
    input={
        "startUrls": [{"url": "https://www.amazon.com/s?k=laptops"}],
        "maxItems": 100,
        "includeReviews": True,
    }
)

# Get and analyze products
products = get_dataset(run_id=run_result.run_id)
avg_price = sum(p.price for p in products.items if p.price) / len([p for p in products.items if p.price])
print(f"Average price: ${avg_price:.2f}")

# Find best rated
best = max(products.items, key=lambda p: p.rating or 0)
print(f"Best rated: {best.title} ({best.rating} stars)")

Batch Scraping

# Scrape multiple websites
websites = [
    {"url": "https://example.com/products"},
    {"url": "https://example.org/items"},
    {"url": "https://example.net/catalog"},
]

for site in websites:
    run_result = run_actor(
        actor_id="apify/web-scraper",
        input={
            "startUrls": site,
            "selector": ".product",
            "extractData": {"title": "h2", "price": ".price"},
        }
    )
    data = get_dataset(run_id=run_result.run_id)
    print(f"{site['url']}: {len(data.items)} items found")

Popular Actors

Actor	Purpose
google-maps-scraper	Extract business listings from Google Maps
instagram-scraper	Scrape Instagram profiles and posts
tiktok-scraper	Scrape TikTok videos and profiles
amazon-scraper	Extract Amazon product data
youtube-scraper	Scrape YouTube videos and channels
web-scraper	Generic website scraper with custom selectors
serp-scraper	Scrape search engine results

Pros

• ✅ 2,500+ Pre-Built Scrapers: Ready-to-use for popular sites
• ✅ No Scraping Code: Use battle-tested actors
• ✅ Proxy Management: Built-in proxy rotation
• ✅ Anti-Bot Bypass: Handles CAPTCHAs and bot detection
• ✅ Scalable: Cloud infrastructure for large jobs
• ✅ Multiple Export Formats: JSON, CSV, Excel
• ✅ Scheduling: Run scrapers on a schedule

Cons

• ❌ API Token Required: Need Apify account
• ❌ Cost: Free tier has limits, paid for heavy use
• ❌ Platform Dependency: Tied to Apify infrastructure
• ❌ Rate Limits: API rate limits apply

When to Use

Use Apify MCP when:

• You need to scrape popular websites without custom code
• You need reliable scraping with proxy rotation
• You're scraping at scale (hundreds/thousands of pages)
• You need to bypass anti-bot measures
• You want scheduled recurring scrapes

Avoid Apify MCP when:

• You can write custom scrapers easily
• You're on a tight budget (free tier limits)
• You need to scrape niche websites without pre-built actors
• You need real-time scraping (cloud latency)

Resources

• Apify Documentation
• Actor Store
• GitHub Repository
• Apify Console