Website Metadata Extractor

What does Website Metadata Extractor do?

Website Metadata Extractor reads the HTML head of public web pages and saves structured metadata for SEO audits, canonical link checks, and social preview reviews.

What website data can you extract?

• url and originUrl identify the final page URL and its starting URL.
• title and lang contain the page title and declared HTML language.
• meta contains named meta tags, including descriptions, robots directives, Open Graph tags, and X Card tags when present. X Card tags use the twitter: prefix in HTML.
• links groups head links by their rel value, such as canonical or alternate.
• statusCode and extractedAt record the HTTP response status and extraction time.

How to use Website Metadata Extractor

1. Enter one or more public HTTP or HTTPS URLs in urls.
2. Run the Actor. Direct connections are enabled by default.
3. Open the run dataset to review results. Export JSON to preserve the nested metadata and link fields. Flatten those fields before converting to CSV for a spreadsheet.

{
  "urls": ["https://example.com/"],
  "proxy": { "useApifyProxy": false },
  "pagination": { "enabled": false }
}

Example dataset row

{
  "url": "https://example.com/",
  "originUrl": "https://example.com/",
  "title": "Example Domain",
  "lang": "en",
  "meta": { "viewport": "width=device-width,initial-scale=1" },
  "links": { "icon": ["data:,"] },
  "statusCode": 200,
  "extractedAt": "2026-11-10T06:14:51.380Z"
}

Crawl settings and coverage

• Supply up to 1,000 distinct starting URLs. URL fragments are stripped for deduplication.
• Enable pagination.enabled to follow matching links on the same hostname. Set pagination.linkSelectors and pagination.maxDepth to control discovery.
• Use rateLimits and request to tune concurrency, requests per minute, timeouts, and retries.
• The Actor reads the HTML returned by the server. Metadata added only by browser-side JavaScript and pages behind sign-in or bot challenges may be unavailable.
• Missing tags are omitted. After retries are exhausted, a failed request produces an error row with the URL, origin URL, and timestamp. Check for error rows before using the dataset.