SEO Audit Tool: Website SEO Checker, 0-100 Score, $5/1K Pages

SEO Audit Tool: Website SEO Checker, 0-100 Score

This SEO audit tool crawls a website over plain HTTP and gives every page a score from 0 to 100, with a list of what is wrong and how to fix it. It also adds one summary row for the whole site. There is no browser involved, so a small site is done in a minute or two.

It follows robots.txt, stays on the domain you start from, and sends a few requests at a time so it does not load the site.

How to run an SEO audit, step by step

1. Open the Actor and paste the site address into Start URLs, for example https://www.example.com/. The prefilled address is a public practice site, so you can try a run before using your own.
2. Set the Page limit. Ten pages is a quick test. Fifty is the default. You can go up to 10,000.
3. Optional: lower Link depth to stay near the home page, or add Excluded URL patterns such as /cart to skip sections.
4. Click Start. When the run finishes, open the Overview table. Sort by score to see the weakest pages first.
5. Open the Issues view to read each problem with its fix. The last row of the dataset has type: "site" and holds the site score and the most common issues.

An empty or invalid start URL ends the run with a message instead of an error.

What it checks

Each distinct issue on a page costs points, and a page starts at 100. The same issue is only counted once per page. The score never goes below 0. Pages that return an HTTP error score 0.

Severity	Points off	Issues
Critical	15	HTTP error (4xx or 5xx), noindex on the page, missing title
High	8	Missing H1, missing meta description, broken internal links, canonical pointing elsewhere
Medium	4	Title not 30 to 60 characters, meta description not 70 to 160, more than one H1, duplicate title, duplicate meta description
Low	2	Images without an alt attribute, no structured data, no Open Graph title or image

Duplicates are found across the pages of one run. The first page with a title or description keeps it, and later pages with the same text point back to it in duplicateTitleOf or duplicateMetaOf.

The site row also reports whether robots.txt exists, whether a sitemap exists and how many URLs it lists, whether plain http redirects to https, whether /llms.txt exists, the site score (the average of the page scores) and the ten most common issues.

How much it costs

You pay per audited page: $5 per 1,000 pages on the Free plan ($0.005 each, lower on paid plans). A page is charged only after its row is saved. Pages that fail, non-HTML files such as PDFs, and the site summary row are free. If you set a maximum charge for the run, the crawl stops when it is reached.

Example: a 200-page site costs about $1.00. A run with the default 50-page limit costs at most $0.25. If you turn on Issues only, pages with no issues are not saved and not charged.

Input example

{
  "startUrls": [{ "url": "https://books.toscrape.com/" }],
  "maxPages": 10,
  "maxDepth": 2,
  "excludeUrlPatterns": ["/cart"],
  "maxConcurrency": 4
}

Other options: includeSubdomains, checkExternalLinks (sends HEAD requests to outbound links, up to 25 per page, and respects those sites' robots.txt) and outputIssuesOnly.

Output example

This is a real row from a test run against a small local site (URLs shortened, issue list trimmed to three of the seven found):

{
  "type": "page",
  "url": "http://127.0.0.1:8777/notebooks.html",
  "statusCode": 200,
  "depth": 1,
  "title": "Notebooks",
  "titleLength": 9,
  "metaDescriptionLength": 123,
  "h1Count": 2,
  "wordCount": 3,
  "imagesTotal": 2,
  "imagesMissingAlt": 1,
  "schemaTypes": [],
  "duplicateTitleOf": "http://127.0.0.1:8777/journals.html",
  "issues": [
    { "code": "TITLE_LENGTH", "severity": "medium", "message": "The title is 9 characters (aim for 30 to 60).", "fix": "Rewrite the title so it fits roughly 30 to 60 characters." },
    { "code": "MULTIPLE_H1", "severity": "medium", "message": "The page has 2 H1 headings.", "fix": "Keep one H1 and turn the others into H2 or lower." },
    { "code": "DUPLICATE_TITLE", "severity": "medium", "message": "Same title as http://127.0.0.1:8777/journals.html.", "fix": "Give each page its own title." }
  ],
  "score": 78
}

The site row from the same run:

{
  "type": "site",
  "siteScore": 52,
  "pagesCrawled": 6,
  "pagesAudited": 4,
  "pagesFailed": 2,
  "robotsTxtFound": true,
  "sitemapFound": true,
  "sitemapUrlCount": 2,
  "llmsTxtFound": false,
  "topIssues": [{ "code": "NO_SCHEMA", "severity": "low", "pages": 3 }, { "code": "NO_OPEN_GRAPH", "severity": "low", "pages": 3 }],
  "brokenLinkTargets": [{ "url": "http://127.0.0.1:8777/old-page.html", "status": 404, "foundOn": ["http://127.0.0.1:8777/"] }]
}

Bulk lists

Use a long list of start URLs from an Apify Dataset or from a file. Pick the Dataset under Dataset with the list and write the name of its field under Field in the dataset. Leave the field name empty to use the first text field. Or paste the web address of a .txt, .csv or .json file under List file URL. A .csv file uses its first column. A header row is skipped. The file can be up to 10 MB. Entries from the Dataset and the file are added to the ones typed under Start URLs, and duplicates are removed. All entries are start points of the same crawl, and the run gives one site summary row. The price per audited page does not change.

Related Actors

• Tech Stack Detector: Technologies, email provider and DNS for any website.
• Domain Authority Checker: Bulk domain authority scores for websites.
• Sitemap Extractor: All URLs from a sitemap.
• Broken Link Checker: Find broken links on a site.
• AI Crawler Access Audit: Which AI crawlers a site allows in robots.txt.

FAQ

Why do the scores differ from Lighthouse or other SEO tools? Every tool picks its own checks and weights. This one scores only what it can read from the HTML and headers, and the weights are in the table above. It does not measure speed, Core Web Vitals or backlinks, so use it for on-page and crawl problems, not as a full ranking forecast.

Does it respect robots.txt? Always, and you cannot turn it off. Disallowed URLs are skipped and counted in skippedByRobots. Wildcard rules such as Disallow: /*.pdf$ and a Crawl-delay are honoured. Retry-After headers are honoured too. If the site keeps answering 403 or 429, the run stops early and says so in stoppedReason. It does not rotate proxies or hide who it is. The user agent is SEOAuditToolBot.

Does it run JavaScript? No. It reads the HTML the server sends. Pages that build their content in the browser may show as thin or missing a title.

How are broken internal links and duplicates found? The crawl keeps every page result in memory until it has finished. Then it fills in brokenLinks from the status of every URL it fetched, marks duplicate titles and descriptions, recalculates the score, and only then saves and charges each page row. That is why rows appear in the dataset after the crawl, not one by one during it. If you set a maximum charge, saving stops when it is reached and the remaining pages are neither saved nor charged. A page is "broken" only if the link target was fetched in the same run, so links to pages beyond your page limit or depth are not checked.

What are the limits? Up to 10,000 pages per run, depth up to 20, and 10 requests at a time per host (4 by default). Pages larger than about 2 MB are cut off. Only the domain of your start URL is crawled, plus its subdomains if you turn that on.