Extracteur de Pages Web

Pull the title, metadata, headings, links, images and text out of any public page

100 % gratuit. Cet outil a besoin de notre serveur pour effectuer le travail : ce que vous soumettez y est traité puis supprimé, il n'est jamais conservé.

Selector supports a tag name, .class or #id. Full CSS selectors are not supported - anything more complex returns nothing rather than a wrong answer.

12 of 12 scrapes left today.

Jump to the next useful tool without leaving your flow.

See more

À propos de cet outil

Pull the structure out of any public web page. Give it a URL and it returns the title, meta description, canonical, language, Open Graph tags, every heading in order, all links marked internal or external and flagged for nofollow, every image with its alt text and declared size, and the readable text with a word count. An optional simple selector - a tag name, a class or an id - returns just the text of matching elements, which is enough for pulling prices, titles or list items without writing a scraper. Links and text can be copied out in one click.

Flux de travail

Auditer une page avant de la publier

  1. 1Scraper la pageRécupérez ses titres, liens, images et métadonnées.Vous êtes ici
  2. 2Vérifier les balises metaTitre et description tels que les moteurs de recherche les lisent.Ouvrir
  3. 3Prévisualiser la carte de partageVoyez à quoi ça ressemble une fois publié sur les réseaux sociaux.Ouvrir
  4. 4La capturerConservez une capture de la page finalisée pour vos archives.Ouvrir

Comment utiliser cet outil

  1. Paste a URLAny public http or https page. It is fetched server-side, so pages that block browser requests still work.
  2. Add a selector if you want oneA tag name, .class or #id narrows the result down to just those elements.
  3. Browse the tabsOverview, headings, links, images, meta and text each have their own view, with copy buttons for the lists.

Pourquoi l’utiliser

  • Everything in one pass - metadata, headings, links, images and text - rather than one tool per field.
  • Links are deduplicated and marked internal or external, with nofollow flagged.
  • Images include alt text, so missing alt attributes are immediately visible.
  • Open Graph and meta tags are listed together for checking how a page will share.
  • Scripts and styles are stripped before text extraction, so the word count reflects real content.
  • Copy buttons for the link list and the page text.

Usages courants

  • Auditing a page for missing alt text, headings or meta description.
  • Exporting every outbound link on a page for a backlink or compliance review.
  • Checking how a page will look when shared, by reading its Open Graph tags.
  • Pulling a list of product titles or prices using a class selector.
  • Counting real word content on a competitor page.
  • Grabbing the readable text of an article for analysis.

Conseils pour de meilleurs résultats

  • Start with the Overview tab - most SEO problems are visible there in seconds.
  • Use the internal and external filters on the links tab before copying.
  • Turn off Extract page text for a much smaller, faster result when you only want links.
  • A class selector matches any element carrying that class, so keep it specific.
  • The heading tab shows nesting by indentation, which makes a broken heading order obvious.
  • Some sites serve different markup to servers than to browsers, so compare against the live page if something looks off.

Erreurs à éviter

  • Expecting full CSS selector support - only a tag name, .class or #id are handled.
  • Scraping a page whose content is rendered entirely by JavaScript after load, which returns the empty shell.
  • Reading a login or consent wall as the real page.
  • Assuming the link list is a full site crawl - it covers the one page you fetched.
  • Using scraped content in a way the site prohibits; check the terms and robots rules first.

Questions fréquentes

Title, meta description, canonical, language, Open Graph and meta tags, every heading with its level, deduplicated links marked internal or external and flagged for nofollow, images with alt text and declared dimensions, and the readable page text with a word count.

Only simple ones: a tag name, .class or #id. Anything more complex returns nothing rather than a wrong answer, which is the honest behaviour for a regex-based extractor.

Because the page renders its content with JavaScript after load. This tool fetches the HTML the server sends, so a client-rendered page returns its empty shell. For those, use the Website Screenshot tool, which runs a real browser.

No. The fetch is anonymous with no cookies or session, so anything behind a login returns the signed-out page.

That depends on the site. Many publish terms or a robots file describing what automated access they permit, and copyright still applies to the content you retrieve. Check before you use scraped material, particularly commercially.

Related guides