HTML Metadata Inspector

Read title, meta, canonical-link and base declarations from pasted HTML without visiting a website.

Paste HTML source. The detached fragment is never rendered; links stay as text. Maximum 100,000 UTF-16 input units.

Processing stays in this browser. Source scripts and resources are not executed or fetched.

Read metadata declarations from HTML source

Paste HTML and choose Inspect HTML. The report lists title elements, meta elements, canonical link elements and base href declarations in parsed order. Each row identifies the element kind, a name and its text or attribute content. The tool does not contact a URL or inspect HTTP response headers.

Example: title, description and sharing title

Load the example to obtain four rows: title Example page, description A small example, og:title Shared title and canonical https://example.com/page. The canonical value remains literal text. It is not followed, normalized, validated against the page address or submitted to a search engine.

Interpreting duplicate and incomplete declarations

Repeated declarations remain separate, helping you notice conflicting values without guessing which a consumer will use. A meta element's name is taken from name, then property, then http-equiv, or charset when present. Charset content comes from its charset attribute; other meta content comes from content. Missing attributes remain blank. Malformed declarations containing several competing identifying attributes need manual review.

Canonical detection is case-insensitive within the rel token list. Other link types are excluded. Base href appears as a separate row but is not applied when reading other URLs. Title text has whitespace collapsed. JSON-LD scripts, html language attributes, robots.txt, sitemap entries, runtime changes and server headers are outside this report. A populated report cannot establish indexing eligibility or ranking.

Fragment parsing and export

A detached template holds the parsed source and is never attached to the live page. Its parser ignores html, head and body wrapper tags; consequently this inspector does not verify that metadata is located in a valid head. Nested template contents are not traversed. Scripts and resources do not execute or load. See the browser's template parsing notes for the fragment behavior.

The table previews 200 matches; JSON copy and download contain all rows. Limits are 100,000 UTF-16 input units, 10,000 parsed elements and 2,000 report rows, with a conservative two-million-character report budget. Editing clears stale output. Use HTML Link Inspector for anchor destinations and HTML Heading Outline Viewer for the document's heading sequence.