No-Code CSS Selector Scraper - Fast Cheerio HTTP Crawler

Scrape static websites without writing code: map field names to CSS selectors, point the Actor at your start URLs, and get clean JSON, CSV, or Excel. Page title, meta description, canonical URL, H1s, Open Graph image, and JSON-LD structured data are extracted automatically. Need full control? Add an optional JavaScript page function.

What can this CSS selector scraper do?

• No-code extraction rules - each rule has a name, a CSS selector, a type (text, html, or attribute), and an optional multiple flag for lists.
• Automatic page metadata - title, description, canonical URL, language, H1 headings, and Open Graph image.
• Automatic JSON-LD - parses application/ld+json structured data (products, articles, organizations, events...).
• Recursive crawling - follow links matched by a linkSelector, filtered by include globs and excludes, with depth and page limits.
• JavaScript page function - return any JSON from each page when CSS rules are not enough; customData is available in context.
• Production controls - concurrency, retries, timeouts, custom headers, Apify Proxy or your own proxies, and robots.txt compliance (on by default).
• Clean output - empty fields can be removed automatically.

What data can I extract?

Field	Description
url	The page URL
your rule names	Values from your CSS rules (string, or array when multiple is on)
title, description, canonicalUrl, language, h1, image (Open Graph)	Page metadata (when includePageMetadata is on)
jsonLd	Parsed JSON-LD objects (when includeJsonLd is on)

When a page function is supplied, its return value replaces the CSS-rule output.

How much does it cost?

Pay-per-event: $0.001 per result ($1.00 per 1,000 records) plus a tiny Actor-start fee. maxResultsPerCrawl caps the number of billable records. Check the Pricing tab for current rates.

How to scrape a website with CSS selectors

1. Add start URLs.
2. Add extraction rules, for example {"name": "price", "selector": ".price_color", "type": "text"}.
3. Optional: set linkSelector (e.g. a[href]) and globs to crawl more pages.
4. Click Start and export the dataset.

Input example

{
  "startUrls": [{ "url": "https://books.toscrape.com/" }],
  "extractionRules": [
    { "name": "titles", "selector": "article.product_pod h3 a", "type": "attribute", "attribute": "title", "multiple": true },
    { "name": "prices", "selector": "article.product_pod .price_color", "type": "text", "multiple": true }
  ],
  "linkSelector": "li.next a",
  "maxCrawlingDepth": 3,
  "maxPagesPerCrawl": 50
}

Output example

{
  "url": "https://books.toscrape.com/",
  "titles": ["A Light in the Attic", "Tipping the Velvet"],
  "prices": ["£51.77", "£53.74"],
  "title": "All products | Books to Scrape - Sandbox"
}

Use cases

• Product catalogs and prices from static ecommerce sites
• Directory and listing pages
• SEO metadata and JSON-LD collection across many URLs
• Quick one-off data pulls without writing a crawler

API example

from apify_client import ApifyClient

client = ApifyClient("<YOUR_APIFY_TOKEN>")
run = client.actor("groupoject/cheerio-web-scraper").call(run_input={
    "startUrls": [{"url": "https://books.toscrape.com/"}],
    "extractionRules": [{"name": "heading", "selector": "h1", "type": "text"}],
})
print(list(client.dataset(run["defaultDatasetId"]).iterate_items())[:3])

Use with AI agents (MCP)

Connect the Apify MCP server at https://mcp.apify.com?tools=groupoject/cheerio-web-scraper in Claude Desktop, Cursor, VS Code or any MCP client, and your AI assistant can call this actor as a tool. Example prompt: "Scrape the product names and prices from these 20 category URLs with these CSS selectors and return a table."

FAQ

Does it run JavaScript on the page? No. It fetches raw HTML over HTTP, which makes it fast and cheap but unsuitable for sites that render content in the browser.

How is it different from Apify's Cheerio Scraper? Apify's Cheerio Scraper is code-first (you write a page function). This Actor adds no-code CSS rules plus automatic metadata and JSON-LD, while still accepting a page function.

Does it respect robots.txt? Yes, by default (respectRobotsTxtFile).

Related Actors

• Website to Markdown Crawler - full-page Markdown for LLMs.
• Technical SEO Audit Crawler - SEO issues per page.