Website Metadata Extractor: Open Graph, Twitter Card & SEO Tags

Give it a list of URLs and get one clean row per page with everything a link preview, an SEO audit or a data pipeline needs: title, meta description, canonical URL, Open Graph tags (og:title, og:description, og:image, og:type, og:site_name), Twitter Card tags, favicon and Apple touch icon, page language, robots meta and X-Robots-Tag, hreflang alternates, RSS/Atom feeds, schema.org JSON-LD types (and the raw JSON-LD if you want it), H1s, heading counts, images without alt text and a short list of SEO issues.

No browser, no login: it reads the public HTML of each page, respects robots.txt and is fast and cheap.

What you can use it for

• Link previews: build rich cards (title, description, image, favicon) for any URL in your app, newsletter or CRM
• Quick SEO audits: check thousands of pages for missing titles, long descriptions, missing og:image, missing canonical, several H1s or a noindex tag
• Social sharing QA: make sure every page you share has the right Open Graph and Twitter Card image before a campaign
• Data enrichment: turn a list of company websites into site name, description, logo/icon, language and social image
• Structured data checks: see which pages carry Product, Article, Organization, FAQPage, BreadcrumbList or Recipe JSON-LD
• Migration checks: confirm canonicals, hreflang and robots tags after a redesign or domain move

Why this one

• $1 per 1,000 URLs, $0.001 per run start. The best-known metadata actor charges about five times as much.
• All the tags in one row: Open Graph, Twitter Card, icons, canonical, hreflang, feeds and JSON-LD together
• SEO issues spelled out in plain words on every row, ready to filter in a spreadsheet
• Polite and safe: honours robots.txt by default, one request per page, clear error column when a site refuses automated access
• Fast: up to 20 pages in parallel, 256 MB memory

Input

Field	What it does
urls	Page URLs or bare domains, one per line
includeJsonLd	Add the page's raw schema.org JSON-LD blocks
respectRobotsTxt	Skip URLs disallowed by robots.txt (default on)
skipFailed	Leave out pages that could not be fetched, so you pay only for successes
concurrency	Pages fetched at the same time (default 8)
maxResults	Hard cap on URLs processed

Example:

{
  "urls": ["https://www.apple.com", "https://github.com", "https://www.python.org/about/"],
  "includeJsonLd": false
}

Output

{
  "url": "https://www.python.org/about/",
  "finalUrl": "https://www.python.org/about/",
  "status": 200,
  "title": "About Python™ | Python.org",
  "titleLength": 26,
  "description": "The official home of the Python Programming Language",
  "canonical": null,
  "language": "en",
  "ogTitle": "About Python™",
  "ogImage": "https://www.python.org/static/opengraph-icon-200x200.png",
  "ogType": "website",
  "ogSiteName": "Python.org",
  "twitterCard": null,
  "favicon": "https://www.python.org/static/favicon.ico",
  "appleTouchIcon": "https://www.python.org/static/apple-touch-icon-144x144-precomposed.png",
  "feeds": ["https://www.python.org/blogs/feed/"],
  "hreflang": [],
  "jsonLdTypes": ["WebSite", "SearchAction"],
  "h1": ["About"],
  "headingCounts": {"h1": 1, "h2": 4},
  "images": 3,
  "imagesWithoutAlt": 1,
  "issues": ["missing canonical"]
}

Other columns: descriptionLength, robotsMeta, xRobotsTag, ogDescription, ogUrl, ogLocale, twitterSite, twitterTitle, twitterImage, themeColor, author, publishedTime, modifiedTime, keywords, generator, viewport, and jsonLd when asked for. A page that could not be read has an error column (for example HTTP 403 (site blocked automated access) or blocked by robots.txt).

How much does it cost?

$1.00 per 1,000 URLs on the Free plan plus $0.001 per run, with automatic discounts on paid Apify plans. Platform compute is included. Turn on skipFailed to pay only for pages that returned metadata.

Worked examples:

• Link previews for 200 URLs a day from your app: $0.20 a day
• SEO check of a 5,000-page sitemap: $5.00
• Enrich 1,000 company homepages with site name, description and icon: $1.00

FAQ

Is this legal? It reads the same public HTML any browser or social network preview bot reads, one request per page, and skips anything robots.txt disallows. No logins, no paywalls, no personal data.

Why did some sites return an error? A few large sites refuse automated requests (HTTP 403 or a bot check). They appear with an error column instead of wrong data, and with skipFailed on they cost nothing.

Does it run JavaScript? No. Open Graph, Twitter Card, canonical and JSON-LD tags are almost always in the server HTML because social networks and search engines need them there. Pages that build their tags only in the browser will show them as missing.

Can I crawl a whole site? Give it the URLs. To get every URL of a site first, run a sitemap extractor and pass its output here.

What counts as an SEO issue? Missing title, title over 60 characters, missing meta description, description over 160 characters, missing og:title, missing og:image, missing canonical, no H1 or more than one H1, and a noindex robots tag.

Tips

• Feeding a sitemap: run a sitemap URL extractor first, then pass its url column here to audit every page of a site in one go.
• Monitoring: schedule a weekly run on your key landing pages and send rows whose issues list is not empty to Slack, so a broken social image or an accidental noindex is caught the same week.
• Clean previews: prefer ogImage for the card image, fall back to twitterImage, then appleTouchIcon; use ogSiteName or the domain as the source label.
• Large lists: keep concurrency at 8 to 12; lists that mix many different sites finish fastest.

Integrations

Works with Google Sheets, Slack, Zapier, Make, Airtable, webhooks, the Apify API and the Python and JavaScript clients. Download as JSON, CSV, Excel, XML or HTML.

Changelog

• 2026-10-20 — First release: title, description, canonical, Open Graph, Twitter Card, icons, language, robots, hreflang, feeds, JSON-LD types and raw JSON-LD, headings, image alt count and SEO issues.