Website Metadata Extractor

How do I extract ALL of a page’s metadata — meta tags, Open Graph, canonical, favicons, feeds — without uploading the page?

Open AnyTool’s Website Metadata Extractor, open your page and copy its HTML source (View Source / Ctrl+U), and paste it in. It pulls out EVERY piece of page metadata a browser or crawler reads and displays it grouped and copyable: General (title, description, <html lang>, charset, viewport, author, robots, generator, theme-color, keywords, application-name and more), Social (every og:* and twitter:* tag plus a resolved Facebook/LinkedIn-style share-card preview), Links & Icons (canonical, rel=alternate hreflang language variants, RSS/Atom/JSON feeds, every favicon / apple-touch-icon / mask-icon, the PWA manifest, and preconnect / dns-prefetch / preload hints), Structured Data (the list of JSON-LD @types present), and Other meta (anything else). Every value has a one-click copy and the whole result exports as JSON or a flat key→value text list. It’s an EXTRACTOR (a developer/data tool), not an SEO audit — for pass/warn recommendations use the sibling Meta Tag Analyzer. Everything runs in your browser, so your page HTML is never uploaded — safe for staging. Your browser can’t fetch most live URLs (CORS) and there’s no server proxy, so paste the source; it reads the markup, it doesn’t render the page or fetch the actual favicon / image files.

Paste your page’s HTML source — every meta tag, link and JSON-LD type is extracted and grouped
Groups: General, Social (og + twitter + share-card preview), Links & Icons, Structured Data, Other meta
Extracts canonical, hreflang alternates, RSS/Atom/JSON feeds, all favicons, manifest, resource hints
Per-value copy + full JSON export (or flat key→value text) — a developer/data tool, not an audit
100% client-side — page HTML never uploaded; safe for staging / unreleased pages

What is

Page metadata extraction

Page metadata is the machine-readable information a web page carries in its markup for browsers, search engines and social platforms — separate from the visible body content. It includes the <title>, the <meta> tags (description, keywords, author, robots, viewport, charset, generator, theme-color and more), the Open Graph (og:*) and Twitter Card (twitter:*) social tags, the <html lang> language, and the <link> relations: canonical, rel=alternate hreflang language/regional variants, RSS/Atom feeds, favicons and apple-touch-icons, the PWA manifest, and performance hints (preconnect, preload). Structured data (JSON-LD) declares schema.org @types. Extracting metadata means parsing all of it out of the HTML and presenting it as clean, structured, exportable data — distinct from auditing it (scoring meta tags against SEO best practice).

SEO & Web

Related terms

Meta tagsOpen GraphTwitter CardCanonical URLhreflangFaviconRSS/Atom feedJSON-LDWeb app manifest

Frequently Asked Questions

They are siblings for two different jobs. The Website Metadata Extractor is a developer/data tool: paste your page HTML and it inventories every piece of metadata — the title, every meta tag, all Open Graph and Twitter tags (with a resolved share-card preview), the canonical, hreflang alternates, feeds, favicons, manifest, resource hints and JSON-LD @types — grouped into a clean key→value view you can copy per value or export as JSON. It does not score or judge anything. The Meta Tag Analyzer is an SEO audit: it checks your title/description length, missing og:image, non-absolute URLs and the like, and returns pass / warn findings with recommendations. Use the Extractor to see and export exactly what metadata a page carries; use the Analyzer to grade and improve it. Both run 100% in your browser.

The Extractor pulls out and displays ALL your page metadata as structured, exportable data — a developer/data tool. The Meta Tag Analyzer audits your meta tags with pass/warn checks and SEO recommendations.

It groups everything a page carries. General: <title>, meta description, <html lang>, charset, viewport, author, robots, generator, theme-color, keywords, application-name, publisher, referrer, color-scheme and other name= meta. Social: every og:* and twitter:* tag, plus a resolved Facebook/LinkedIn-style share card (title, description, image, domain) built with the correct <title> and meta-description fallbacks. Links & Icons: the rel=canonical, rel=alternate hreflang language variants, RSS / Atom / JSON feeds, every icon / apple-touch-icon / mask-icon (with sizes and type), the rel=manifest PWA manifest, and preconnect / dns-prefetch / preload / prefetch hints. Structured Data: the distinct schema.org @types from the JSON-LD blocks. Other meta: any remaining name= / property= / http-equiv= tags. Everything is copyable per value and exports as JSON.

The title, every meta tag, Open Graph and Twitter tags (with a share preview), canonical, hreflang alternates, RSS/Atom/JSON feeds, all favicons, the manifest, resource hints, and the JSON-LD @types.

Because the tool is 100% client-side with no server proxy — which is what keeps your page private — your browser can only fetch pages that explicitly allow cross-origin requests (CORS) or are on the same origin. Almost all live sites block that, so the reliable path is to open your page, use View Source (Ctrl+U), copy the HTML and paste it in. The optional “try to fetch” box is there for the rare CORS-enabled case; when it fails, it tells you to paste the source instead. And it reads what is in the markup — it does not render the page or fetch external resources, so it shows the og:image / favicon / manifest URLs, not the actual image or file. Nothing you paste is ever uploaded.

Paste the HTML source. Your browser can’t fetch most live URLs (CORS) and there’s no server proxy, so the URL box only works for CORS-enabled or same-origin pages.

The Website Metadata Extractor is 100% client-side. The HTML you paste is parsed with the browser’s DOMParser and everything is displayed and exported locally; nothing is fetched, uploaded, logged or stored on a server, and no page is crawled (there is deliberately no server proxy). Unlike online metadata extractors that send your whole page — or a URL they crawl server-side — to a remote service, this is safe for staging, unreleased or internal pages you don’t want to expose before launch. It works offline once cached, and closing the tab discards everything you entered.

No. The HTML you paste is parsed entirely in your browser with the built-in DOM parser — nothing is uploaded, logged or stored, so it’s safe for staging pages.

Paste a page’s HTML source and extract ALL of its metadata — every meta tag, Open Graph and Twitter tag (with a resolved share preview), canonical, hreflang alternates, RSS/Atom/JSON feeds, all favicons, the manifest, resource hints and JSON-LD @types — grouped, copyable and exportable as JSON. A developer/data extractor, not an SEO audit. 100% client-side; your page HTML is never uploaded.

Features

Extracts every meta tag, link and JSON-LD @type from pasted page HTML, grouped into 5 clean sections
Social group: all og:* and twitter:* tags + a resolved Facebook/LinkedIn-style share-card preview
Links & Icons: canonical, hreflang alternates, RSS/Atom/JSON feeds, all favicons, manifest, resource hints
Per-value copy + full JSON / flat text export — a developer/data tool, distinct from the Meta Tag Analyzer audit
100% client-side — page HTML never uploaded; honest CORS-paste-source / reads-markup-not-rendered note

PriceFree

CategorySEO & Web

PlatformAll (browser-based)

PrivacyPage HTML is parsed locally with DOMParser and never uploaded; no page is crawled (no server proxy)

Input: Paste your page’s full HTML source (View Source / Ctrl+U), Paste just the <head> fragment, Optionally fetch a CORS-enabled / same-origin URL

Output: Grouped metadata (General / Social / Links & Icons / Structured Data / Other) with per-value copy, Resolved social share-card preview, Full JSON export or flat key→value text list

Detailed Explanation

How It Works

What the Website Metadata Extractor Pulls Out

The Website Metadata Extractor is an EXTRACTOR (a developer / data tool), not an SEO audit: the user pastes a page’s HTML source (View Source / Ctrl+U) and the tool pulls out EVERY piece of page metadata a browser or crawler reads and displays it in a clean, GROUPED, copy/exportable view, with no scores and no pass/warn recommendations (that is the sibling Meta Tag Analyzer’s job). It groups the metadata into five sections. General: the <title>, meta description, <html lang>, <meta charset>, viewport, author, robots, generator, theme-color, keywords, application-name, publisher and other name= meta (referrer, color-scheme, format-detection, rating and the like). Social: every og:* and twitter:* tag, plus a RESOLVED Facebook / LinkedIn-style share-card preview (title, description, image, domain) built with the correct <title> and meta-description fallbacks. Links & Icons: the rel=canonical, rel=alternate hreflang language / regional variants, RSS / Atom / JSON feeds, every icon / apple-touch-icon / mask-icon (with sizes and type), the rel=manifest PWA manifest, and preconnect / dns-prefetch / preload / prefetch resource hints. Structured Data: the distinct schema.org @types found across the JSON-LD blocks. Other meta: any remaining name= / property= / http-equiv= tags. Every value has a one-click copy, and the whole result exports as a pretty JSON object or a flat key→value text list.

An extractor / data tool — extracts and DISPLAYS all metadata; it does not score or audit it
Five groups: General, Social (og + twitter + share preview), Links & Icons, Structured Data, Other meta
Extracts canonical, hreflang alternates, RSS/Atom/JSON feeds, all favicons, manifest, resource hints, JSON-LD @types
Per-value copy + full JSON export (or flat key→value text) for pipelines, specs, diffs, bug reports
All parsing runs in the browser via DOMParser — no server call, no CDN, no upload

Methodology

How It Parses and Groups the Metadata

The tool is built on a small, pure, well-typed engine, metadataExtractEngine.ts, that REUSES the shared socialMetaEngine.ts — its parseSocialMeta extracts the og:* and twitter:* dictionaries, the <title>, the meta description and the canonical, and its buildShareCardModel resolves the share-card preview with the documented og → <title> / meta-description fallbacks — so the social handling is exactly the same source of truth the Open Graph Checker and Twitter Card Validator use, imported and NOT modified. On top of that, the new engine parses the whole head with the browser’s DOMParser (falling back to a regex scan so it never throws in a non-DOM environment, and so a pasted <head> fragment still works): it walks every <meta> (name=, property=, http-equiv= and charset), reads <html lang>, and walks every <link>, bucketing each rel — canonical, manifest, the icon family (icon / shortcut icon / apple-touch-icon / mask-icon), rel=alternate split into hreflang variants vs RSS/Atom/JSON feeds (by type) vs other alternates, and the resource-hint family (preconnect / dns-prefetch / preload / prefetch). For structured data it parses each <script type="application/ld+json"> block, recursing @graph and arrays to collect the distinct @type values (with a shallow regex fallback for invalid JSON). It assembles the grouped object plus a flattened, ordered key→value list, and serializes to JSON or text. There is no new dependency and no network.

Reuses socialMetaEngine.ts (parseSocialMeta + buildShareCardModel) for og/twitter + the share preview — engine unchanged
New engine metadataExtractEngine.ts parses the head with DOMParser (regex fallback; never throws)
Links bucketed by rel: canonical, manifest, icons, hreflang alternates, feeds, other alternates, resource hints
JSON-LD @types collected by walking @graph and arrays; shallow regex fallback for invalid JSON
Produces grouped output + a flat key→value list; exports as pretty JSON or text — no new dependency, no network

Limitations

Paste Source (CORS), Reads Markup Not Render, and It’s an Extractor Not an Audit

The tool is explicit about three honest limits. First, paste the page SOURCE: because it is 100% client-side with NO server proxy (which is what keeps your page private), your browser can only fetch CORS-enabled or same-origin pages — almost all live sites block cross-origin requests, so the reliable path is View Source (Ctrl+U) → copy → paste; the optional “try to fetch” box is only for the rare CORS-enabled case and, on failure, tells you to paste instead. Second, it reads the MARKUP, it does NOT render the page or fetch external resources: it extracts the metadata written in the HTML you paste, so it shows the og:image / favicon / manifest URLs (and can attempt to preview an absolute-https og:image), but it does not download the actual favicon image, fetch the linked manifest JSON, or run JavaScript that might inject tags at runtime — what you paste is what it reads. Third, it is an EXTRACTOR, not an SEO AUDIT: it pulls out and displays the metadata as structured, exportable data with no scores or recommendations. For a pass / warn audit of your meta tags with fixes, the sibling Meta Tag Analyzer is the right tool; to validate JSON-LD, Microdata and RDFa structured data, the sibling Structured Data Tester is; to extract the page’s links, the Website Link Extractor is.

Paste your page source — browser can’t fetch most live URLs (CORS); no server proxy keeps it private
Reads the markup; it does not render the page or fetch external resources (shows og:image / favicon URLs, not the files)
JavaScript-injected tags aren’t captured — what you paste is what it reads
It’s an extractor / data tool, not an SEO audit — no scores; for pass/warn use the Meta Tag Analyzer
Siblings: Meta Tag Analyzer (audit), Structured Data Tester (validate JSON-LD), Website Link Extractor (links)

Privacy & Security

Why Extracting Metadata In-Browser Is a Privacy Win

Online metadata extractors send your whole page — or a URL they crawl server-side — to a remote service. The Website Metadata Extractor instead runs entirely on the user’s device: the pasted page HTML is parsed with the browser’s built-in DOMParser, nothing is fetched, uploaded, logged or stored, and no page is crawled (there is deliberately no server proxy). That makes it safe to inspect the metadata of STAGING, UNRELEASED or INTERNAL pages whose source you do not want to expose to a third-party service before launch — a meaningful advantage when auditing or diffing the metadata of a site that is not yet public. The tool works offline once cached as a PWA, and closing the tab discards everything entered. It adds no new dependency: the social handling reuses the small, well-typed socialMetaEngine.ts shared across the SEO & Web family, and the rest of the extraction lives in the compact metadataExtractEngine.ts. Because everything (including the JSON export blob) is built locally, the workflow — paste, extract, copy any value, export JSON — never touches the network.

Page HTML is parsed locally with DOMParser and never uploaded — no server, no CDN, no crawl, no proxy
Safe for staging / unreleased / internal pages, unlike online extractors
Works offline once cached; closing the tab discards your data
No new dependency — reuses socialMetaEngine.ts; the rest is the compact metadataExtractEngine.ts
Dark mode and mobile support with a sticky Extract / Export JSON bar

Page metadata extraction: in-browser (AnyTool) vs online extractors
Capability	AnyTool	Typical online extractors
Processing	Runs entirely in your browser	Sends your page / URL to a server
Your page	HTML never uploaded, works offline	Page or URL crawled server-side
Staging / unreleased pages	Safe — nothing leaves the device	Exposed to a third-party service
Input	Paste your page HTML source (View Source)	Often a URL the tool crawls
Coverage	All meta + og/twitter + canonical/hreflang/feeds/icons/manifest + JSON-LD types	Varies, often og-only
Social preview	Resolved share-card preview from the markup	Sometimes
Export	Per-value copy + full JSON / flat text	Varies
Role	Extractor / data tool (not an audit)	Often mixes in scoring
Cost / signup	Free, no signup	Often gated or ad-heavy

AnyTool extracts a page’s metadata locally from pasted HTML and uploads nothing, which is safe for staging or unreleased pages. It reads the markup (it does not render the page or fetch the actual favicon / image files), and it is an extractor — not an SEO audit (use the Meta Tag Analyzer) or a structured-data validator (use the Structured Data Tester). Comparison as of July 2026.

Cite This Page

Title: Website Metadata Extractor (paste a page’s HTML source to extract ALL of its metadata — every meta tag, Open Graph and Twitter Card tag with a resolved share-card preview, <html lang>, charset, viewport, canonical, hreflang alternates, RSS/Atom/JSON feeds, all favicons and the PWA manifest, resource hints and the JSON-LD @types — grouped into General / Social / Links & Icons / Structured Data / Other meta, with per-value copy and JSON export; a developer/data extractor, not an SEO audit, 100% client-side)
Author: AnyTool
Published: January 1, 2026
Modified: February 8, 2026
Section: SEO & Web
URL: https://www.anytool.tech/website-metadata-extractor

Formatted Citation

AnyTool. "Website Metadata Extractor (paste a page’s HTML source to extract ALL of its metadata — every meta tag, Open Graph and Twitter Card tag with a resolved share-card preview, <html lang>, charset, viewport, canonical, hreflang alternates, RSS/Atom/JSON feeds, all favicons and the PWA manifest, resource hints and the JSON-LD @types — grouped into General / Social / Links & Icons / Structured Data / Other meta, with per-value copy and JSON export; a developer/data extractor, not an SEO audit, 100% client-side)." AnyTool, January 1, 2026. https://www.anytool.tech/website-metadata-extractor

A privacy-first, 100% client-side Website Metadata Extractor that pulls EVERY piece of page metadata a browser or crawler reads out of a pasted HTML source and presents it as clean, grouped, copyable, exportable data, entirely in the browser, with no server call, no CDN, no upload and no server proxy, so the page HTML pasted is never sent anywhere and the tool works offline once cached — making it safe to inspect the metadata of staging, unreleased or internal pages, unlike online extractors that send your whole page or crawl your URL server-side. It is an EXTRACTOR (a developer / data tool), deliberately DISTINCT from the sibling Meta Tag Analyzer, which AUDITS meta tags with pass / warn checks and recommendations: this tool extracts and DISPLAYS the metadata, it does not score or judge it. The user opens their page, copies its HTML source (View Source / Ctrl+U) and pastes it in (an optional “try to fetch a URL” box exists but, with no server proxy, only works for CORS-enabled or same-origin pages — almost all live sites block cross-origin fetches, so the reliable path is pasting the source; a pasted <head> fragment also works). A small, pure, well-typed engine, metadataExtractEngine.ts, REUSES the shared socialMetaEngine.ts (its parseSocialMeta extracts the og:* and twitter:* dictionaries, the <title>, the meta description and the canonical; its buildShareCardModel resolves the Facebook / LinkedIn-style share-card preview — title, description, image, domain — with the documented og → <title> / meta-description fallbacks) — imported and NOT modified — and layers on the rest of the page’s metadata by parsing the head with the browser’s DOMParser (with a regex fallback so it never throws in a non-DOM environment). It groups the result into five sections: GENERAL (the <title>, meta description, <html lang>, <meta charset>, viewport, author, robots, generator, theme-color, keywords, application-name, publisher, and other name= meta such as referrer / color-scheme / format-detection / rating); SOCIAL (every og:* and twitter:* tag plus the resolved share-card preview); LINKS & ICONS (the rel=canonical, rel=alternate hreflang language / regional variants, RSS / Atom / JSON feeds detected by type, every icon / apple-touch-icon / mask-icon with its sizes and type, the rel=manifest PWA manifest, and preconnect / dns-prefetch / preload / prefetch resource hints); STRUCTURED DATA (the distinct schema.org @types collected from every <script type="application/ld+json"> block, recursing @graph and arrays, with a shallow regex fallback for invalid JSON); and OTHER META (any remaining name= / property= / http-equiv= tags). Every value has a one-click copy, and the whole result exports as a pretty JSON object or a flat, ordered key→value text list — ready to drop into a data pipeline, a spec, a diff or a bug report. It is deliberately HONEST about its scope: you paste the source because your browser cannot fetch most live URLs (cross-origin / CORS) and there is no server proxy; it reads the MARKUP and does NOT render the page or fetch external resources (it shows the og:image / favicon / manifest URLs, and can preview an absolute-https og:image, but does not download the actual image files, fetch the linked manifest, or run JavaScript that might inject tags at runtime); and it is an extractor, not an SEO audit — for pass / warn recommendations on your meta tags use the sibling Meta Tag Analyzer, to validate JSON-LD / Microdata / RDFa use the Structured Data Tester, and to extract the page’s links use the Website Link Extractor. It supports dark mode and mobile with a sticky Extract / Export JSON bar, has exactly three ad slots, and adds no new dependency — the social handling reuses socialMetaEngine.ts and the rest is the compact metadataExtractEngine.ts.