Account

What new tool do you want?

Developer Tools

Web Scraper (No-Code)

Extract data from any web page without coding — tables, links, headings, images, meta tags or anything matching a CSS selector — and export to CSV, Excel or JSON.

How to use

  1. 1Enter the page URL.
  2. 2Choose what to extract, or write a CSS selector (and optional attribute).
  3. 3Download the rows as CSV, Excel or JSON.

Frequently asked questions

Can it scrape pages that load data with JavaScript?

No — it reads the HTML the server sends. Data added later by JavaScript won't be there; check the page source (Ctrl+U).

What's a CSS selector?

A pattern that picks elements: h2 for headings, .price for elements with class “price”, table td:nth-child(2) for a table's second column.

Is scraping allowed?

Public facts are generally fine, but respect each site's terms and robots rules, don't overload servers and never collect personal data.

It fetches server-side, on purpose

A browser can't just fetch() an arbitrary third-party page and read its HTML — cross-origin restrictions (CORS) block that by design for almost every site on the internet. So this tool makes the request from the server, not your browser, reads the raw HTML the server actually sent, and parses it there before sending back just the extracted rows. The practical effect: it works on virtually any public page, but it sees the same HTML as curl or 'view source' — nothing that JavaScript adds to the page after load.

Six modes cover most scraping jobs without writing a selector

Links, headings (H1–H6), images, the first table on the page, and title/meta tags are each one click — useful for quick jobs like pulling every link off a page or collecting a site's title and description tags across a list of URLs. The custom CSS selector mode is for everything else: table td:nth-child(2) for one column of a table, .price for every element with that class, or article a for links only inside the main content area. Add an attribute name (href, src, data-id) to pull that attribute's value instead of the element's text.

Worked example: pulling a table from Wikipedia

The tool's default is Wikipedia's 'List of largest cities' page with mode set to 'First table'. Running it fetches the page's HTML, locates the first <table> element, and returns one row per <tr> with one column per header cell — ready to export as CSV, Excel or JSON with the city names, countries and populations intact as plain text, exactly as they appear in the rendered table. Changing the mode to Custom selector with table td:nth-child(1) instead pulls just the first column across every row.

What it can't reach, and the line not to cross

Pages that build their content with JavaScript after the initial load — many single-page apps, infinite-scroll feeds, sites gated behind a 'loading…' spinner — will come back empty, because nothing executes any script; the fix for those is usually to find the underlying API call the page itself makes (visible in the browser's Network tab) and fetch that URL directly, or check if the data exists in the raw page source (Ctrl+U). It's also capped at the first 3 MB of a page and 500 extracted rows, and like any scraper it's bound by the target site's terms of use and robots rules — fine for public facts, not for personal data or anything a site explicitly disallows.

Related tools

More: Developer Tools