Audit any page for on-page SEO issues in seconds — titles, meta tags, headings, canonicals, images, structured data and indexability, scored out of 100.
This tool fetches the URL from our server, so unlike the rest of TweakPad the address you enter does leave your browser. Only publicly reachable pages can be analysed — private addresses and pages behind a login are refused.
About the SEO Analyzer
Fetches a URL and reports what is wrong with it: title and description length measured in rendered pixels rather than characters, heading structure, canonical and robots directives, image alt text, and structured data. Pixel width is the right unit because that is what Google truncates on.
A noindex directive, a 4xx status or a missing title caps the score regardless of what else is right, because each of those stops a page ranking outright and a high score beside one would be misleading. Treat the number as a summary and the findings list as the actual output.
Frequently asked questions
What does this SEO analyzer actually check?
It fetches the URL once and inspects the returned HTML for around thirty on-page signals: the title and meta description (including their rendered pixel width, which is what Google truncates on, not character count), the canonical tag, meta robots directives, the heading hierarchy, the lang attribute, the viewport tag, word count, internal and external link counts, nofollow usage, images missing alt text or width/height attributes, Open Graph and Twitter Card tags, JSON-LD and microdata structured data, HTTPS, redirect chains, server response time, and URL hygiene. Each finding is graded critical, warning, opportunity or notice, and the score subtracts a weight per finding from 100.
Does it crawl my whole site?
No — it audits exactly the one URL you enter. That is a deliberate limit, not an oversight: crawling a site means thousands of requests and minutes of processing, which is a different kind of tool with a different cost. The practical consequence is that checks needing sitewide context are out of scope here, including duplicate titles across pages, orphaned pages with no internal links, and sitewide canonical conflicts. Everything that can be judged from a single page is covered.
Why does this tool send data to a server when the others do not?
Because it has no choice. Every other tool on TweakPad transforms text you already have, so it runs entirely in your browser. An SEO audit needs the HTML of the site being audited, and browsers refuse to fetch a page from another domain — the same-origin policy blocks it outright, and no amount of client-side JavaScript works around that. So the fetch happens on our server. The only thing sent is the URL you type; the page content is analysed in memory and is not stored beyond a short cache of the report.
Why is my score lower than I expected?
A single critical finding costs 30 points, so any page with a noindex tag, a 4xx or 5xx status, or a missing title cannot score above 70. That is intentional — those problems stop a page ranking at all, and a high score beside one would be misleading. Warnings cost 12, opportunities 5 and notices 1. Treat the number as a summary and the findings list as the actual output: a 70 with one fixable critical issue is a much better position than a 70 spread across six warnings.
Can it analyse a JavaScript-rendered page?
Only the HTML the server sends. If your content is rendered client-side by React, Vue or similar, this tool sees the initial HTML shell — so a single-page app may report a missing H1 and thin content that a browser would in fact display. Googlebot does execute JavaScript, so it may see more than this tool does. Server-side rendering or static generation makes your content visible to both, which is why it remains the safer choice for pages that need to rank.
Are there URLs it will refuse to check?
Yes. Only http and https URLs on the standard ports 80 and 443 are accepted, and the target must resolve to a public address — localhost, private network ranges and cloud metadata endpoints are refused, including via a redirect. That is a security requirement rather than a limitation: an endpoint that fetches any address a stranger supplies would otherwise be usable to reach services inside our own network. Pages behind a login also cannot be analysed, since no credentials are sent.