Rendered HTML Checker

Get two snapshots

Open the page you want to check. For the original HTML, use View Page Source and copy the document, or choose Fetch source URL above. For the rendered snapshot in Chrome, open DevTools, select the top-level html element in the Elements panel, then choose Copy > Copy outerHTML. Paste that document into Rendered DOM HTML.

Use the full document, not a selected paragraph or component. Record the final page URL after any redirect. The checker requires the source’s final URL and the rendered snapshot URL to match. A fragment such as #details is ignored for that comparison.

Wait for the content you are investigating to appear, then capture it. Note the browser, capture time, consent choices and what you waited for. A logged-in dashboard and a logged-out page are not equivalent snapshots. Neither are two different product variants that happen to use the same URL. Live source fetching has no browser cookies, so paste the source yourself when matching a particular session matters. Remove private information before submitting either document.

What changed?

Start with the metadata table. It retains all observed head declarations rather than choosing a winner when two titles or canonicals conflict. A missing value means the parser did not find it in that snapshot. It does not prove the live site never supplies it.

Next, open the headings and parsed body text. If the source contains only an app shell but the rendered snapshot contains the article, that is evidence that the supplied snapshots differ. It is not evidence that Google failed to render the article. The comparison does not inspect resource failures, timing, authentication or the JavaScript responsible.

Word-token counts are a quick way to notice substantial changes, not an SEO target. The parser excludes scripts, styles, templates, noscript content and subtrees marked hidden or aria-hidden=”true”. It does not calculate CSS visibility or layout. Text hidden by a stylesheet may still be counted, and languages without word spaces may need a different counting method.

Fix blockers first

A source document with noindex deserves attention before smaller text differences. Google may skip rendering after seeing that directive in the original response. Removing it later with JavaScript is not a dependable way to make a page indexable. Confirm what your server sends and why the directive is there before changing it.

Review canonical changes just as carefully. An initial canonical pointing to one URL and a JavaScript-generated canonical pointing elsewhere create conflicting signals. The report resolves relative canonical URLs against the first supported base declaration, but it cannot tell you which canonical Google selected. Prefer a consistent, intentional declaration and investigate duplicates at their source.

The report also lists inspected metadata found outside the head separately. Do not assume moving a robots tag into the body makes it a valid indexing instruction. Check the actual document structure and response headers. This tool does not include HTTP X-Robots-Tag rules or robots.txt in its verdict, and it deliberately does not issue an overall “indexable” badge.

Check the links

Link changes count HTTP(S) anchor occurrences. If the same destination appears three times in the source and twice in the rendered document, the report records one removed occurrence. It does not claim that the destination disappeared entirely. The full link evidence also retains anchor text and rel attributes, so a label or nofollow change is still available even when the destination stays the same.

Important navigation appearing only after rendering is worth reviewing. An ordinary anchor with a real href gives crawlers a discoverable destination; a click handler alone is a different mechanism. The checker does not click buttons, expand menus or scroll to trigger lazy loading. Capture the state that answers your question and keep that state in the notes.

Images show declared src, srcset and alt values. These are declarations, not proof that an image loaded or that a browser selected a particular srcset candidate. JSON-LD inspection checks JSON syntax and records declared types; it does not validate every schema property or rich-result requirement.

Keep the evidence

Download JSON for the complete report or CSV for the metadata comparison. Send a developer the before-and-after value, page URL and snapshot notes, then capture both documents again after the fix. A narrow example is more useful than a screenshot saying “JavaScript SEO failed.”

No submitted HTML or report is saved by this plugin. Live source mode requests the target page, whose server may log that request. Shared abuse protection uses a salted IP hash, tool identifier and usage date; hosting and verification-provider policies apply separately.

For Google-specific evidence, use Search Console URL Inspection. The HTML in your browser is not necessarily the HTML Google rendered or indexed. A matching comparison does not guarantee crawling, indexing or rankings.

Questions

Reference: Google’s JavaScript SEO guidance and Chrome DevTools DOM documentation. HTML parsing uses the MIT-licensed Masterminds HTML5 parser; it is not a browser renderer.