Skip to main content

Web Crawler API

Web Crawler API

The Web Crawler API enables efficient web scraping and data extraction from websites, optimized for AI model training and automation workflows. It supports various output formats like Markdown and JSON, making it ideal for LLM-friendly data processing. The API is powered by Crawl4AI, designed for high performance and flexibility.

For attribution, please display the CrewAI badge as per the Crawl4AI GitHub repository.

Web Crawler API

Method: POST

URL: {{base_url}}/webcrawl

Authentication

Required Parameters:

Parameter Name

Value

Description

client_id

your_client_id

API authentication ID

Required Headers:

Header Name

Value

Description

client-secret

your_client_secret

API authentication secret

Content-Type

application/json

Data sent in JSON format

Input Formats

JSON Body:

  1. Basic Crawl (Markdown Output, Default)

{
    "url": "https://example.com",
    "deep_crawl": "false"
}

  1. Deep Crawl with BFS Strategy

{
    "url": "https://example.com",
    "deep_crawl": "true",
    "max_pages": 10
}

  1. LLM Extraction with Specific Question

{
    "url": "https://example.com/products",
    "deep_crawl": "false",
    "question": "Extract all product prices"
}

Features of Web Crawler API

  • Blazing Fast Performance: Outperforms many paid services.

  • LLM-Friendly Outputs: Supports JSON, cleaned HTML, and Markdown formats.

  • Multi-URL Crawling: Crawl multiple URLs simultaneously.

  • Media Extraction: Extracts images, audio, and video tags.

  • Link Extraction: Captures internal and external links.

  • Metadata Extraction: Retrieves metadata from web pages.

  • Custom Hooks: Supports authentication, custom headers, and page modifications.

  • User-Agent Customization: Allows custom user-agent settings.

  • Screenshots: Takes screenshots of crawled pages.

  • Custom JavaScript: Executes JavaScript before crawling.

Notes

  • Basic Crawl: Performs a single-page crawl on the provided URL, ignoring max_pages.

  • Deep Crawl: Crawls multiple pages using a BFS (Breadth-First Search) strategy, respecting the max_pages limit.

  • LLM Extraction: Use the question parameter to extract specific data, such as product prices, from the crawled content.

File Support

The Web Crawler API processes web pages and extracts data in the following formats:

  • JSON

  • Markdown

  • Cleaned HTML

Contact

For Base URL, Client ID, Client Secret, or other queries related to the Web Crawler API, please reach out to our support team at:

Email: support@bizdata360.com

Back to top