Website Metadata & SEO Tag Extractor

URLs in, one clean structured object per page out — everything you need for link previews, SEO audits, and social-card debugging, without parsing HTML yourself.

• Core SEO tags — <title>, meta description, canonical, robots, keywords, author, viewport, lang, charset
• Open Graph — the full og:* set (title, description, image, type, url, site_name)
• Twitter Cards — twitter:* (card, title, description, image)
• Favicons — every declared icon, absolute URLs (falls back to /favicon.ico)
• hreflang — language/region alternates as a map
• JSON-LD schema.org — the @types present on the page (Article, Product, Organization, …)
• Headings — the page's h1s
• Honest pricing — one charge per page successfully extracted; fetch failures and non-public hosts are reported and never charged

Input

{
    "urls": ["https://www.apify.com", "https://news.ycombinator.com"],
    "useApifyProxy": false
}

Output (one dataset item per URL)

{
    "url": "https://www.apify.com",
    "final_url": "https://apify.com/",
    "title": "Apify: Full-stack web scraping and data extraction platform",
    "description": "Cloud platform for web scraping…",
    "canonical": "https://apify.com/",
    "og": { "title": "Apify", "image": "https://…/og.png", "type": "website" },
    "twitter": { "card": "summary_large_image" },
    "favicons": ["https://apify.com/favicon.ico"],
    "hreflang": { "en": "https://apify.com/" },
    "schema_types": ["Organization", "WebSite"],
    "robots": "index, follow",
    "http_status": 200
}

Pricing

One event: page — charged per page successfully extracted. A URL that fails to load, or points at a non-public host, is reported with an error and not charged.

Typical uses

• Link-preview cards in apps, chat, and CMSs (OG/Twitter image + title + description)
• SEO audits — canonical, robots, hreflang, schema coverage across a URL list
• Social-card QA — verify what Facebook/X/LinkedIn will render before you publish
• Enrichment — attach structured page metadata to a list of links

Works with AI agents

Exposed via Apify's MCP server — an agent can call "pull the Open Graph and schema types for these URLs" in one step.

FAQ

JS-rendered tags? Metadata (OG/Twitter/canonical/JSON-LD) is almost always in the served HTML, so a plain fetch captures it. If a site is fully client-rendered, enable useApifyProxy or expect fewer tags.

Bulk? Pass many URLs; each is one dataset row and one charge, failures never charged.