Bulk SEO Metadata & Indexability Checker

Audit explicit URLs you own or have permission to check. Spot missing titles, unexpected noindex directives, conflicting canonicals and heading changes, then send consistent page snapshots to n8n, a dashboard or your SEO reporting workflow.

One result per successfully parsed static HTML page. No browser, login, site-wide crawl or AI service.

Get started

1. Paste the full page addresses into Pages to check.
2. Confirm that you have permission to check them.
3. Choose Maximum pages and start the run.
4. Open the Dataset for JSON/CSV, and the run report for failures or limits.

You can try the explicitly permitted scraping sandbox:

{
  "urls": ["https://books.toscrape.com/"],
  "confirmSourceRights": true,
  "maxResults": 1
}

Only the supplied pages are checked. Links and sitemaps are not followed. A maximum of 100 input URLs and 100 page results is supported per run.

What you get

Data	Fields
Page identity	url, finalUrl, httpStatus, checkedAt
Titles and descriptions	title, titleLength, description, descriptionLength
Headings and language	h1, h2, language
Canonical and language alternatives	canonicalUrl, canonicalCount, hreflang
Indexing directives	metaRobots, xRobotsTag, noindexDetected
Social and structured metadata	openGraph, twitter, schemaTypes
Findings	Missing fields, multiple titles/H1/canonicals, noindex and invalid JSON-LD

Missing text is null; lists and objects can be empty. Values are extracted from the received HTML and response headers, not invented. JSON-LD output lists declared types; it is not complete schema validation or rich-result certification.

This shortened example is illustrative, generated from the bundled HTML fixture rather than a live request:

{
  "url": "https://owned.example/office/",
  "finalUrl": "https://owned.example/office/",
  "httpStatus": 200,
  "title": "Example Store | Office Supplies",
  "description": "Office supplies for small teams. Explore our product range.",
  "canonicalUrl": "https://owned.example/office/",
  "h1": ["Office supplies"],
  "h2": ["For small teams"],
  "noindexDetected": false,
  "schemaTypes": ["WebPage"],
  "findings": [],
  "checkedAt": "2026-10-24T00:00:00Z"
}

The complete sample and field definitions ship with the release package.

Understand the results

A valid snapshot: an HTTP 200 HTML document with an HTML element that passes the source checks and is not recognized as an access challenge. Missing SEO fields are useful findings, so a page without a title or description can still be a valid snapshot. Errors, denied requests and non-HTML responses do not become page results. Challenge detection is best-effort and cannot identify every interstitial.

Noindex is a declaration, not a search-engine verdict. noindexDetected flags noindex/none directives conservatively, including engine-scoped headers. It does not show whether Google indexed a page, prove eligibility for indexing or predict rankings. Review the raw directive fields for context.

Duplicates: repeated input URLs are normalized and deduplicated. Different inputs leading to the same final URL produce one snapshot. Canonical declarations do not merge distinct pages. Resolving a redirect alias can still use web requests.

Partial or empty results: inspect RUN_REPORT. It records request counts, errors, stop reason, write uncertainty and skipped prior output attempts. An empty result set is not proof that a site has no SEO metadata. Completed results remain available after an error or time limit; source/output errors finish the run as failed.

Limits and supported pages

• Static HTTP(S) HTML only. No JavaScript rendering, Core Web Vitals, rank tracking, backlink checks or full SEO scores.
• Robots.txt is always respected. Missing robots files (404/410) permit crawling, but do not grant content rights. Other unreadable or failed policies block access.
• Same-host HTTP-to-HTTPS upgrades are supported. Redirects to another hostname, HTTPS downgrades and unusual ports are rejected. Supply the final address when possible.
• Standard ports 80/443 and public destinations only. Private/internal addresses, credentialed URLs, configured source proxies and unsupported compression are rejected.
• The checker requests uncompressed HTML and rejects compressed responses before decoding. Sites that ignore this preference are currently unsupported.
• Requests have an eight-second timeout. Pages are limited to 1 MB, robots files to 128 KB, and each redirect chain to two followed redirects. Result records are limited to 1 MB of JSON.

Public visibility alone is not permission to collect or reuse a website's data. Use the checker for authorized pages and follow the applicable terms.

Advanced limits

Defaults: maxResults: 10, maxSourceRequests: 20, timeoutSeconds: 30. Web request limits include robots, redirects and failures, so the requested page count may not be reached. The processing time limit includes page lookups, requests, parsing and writes; platform startup and cleanup are separate. Larger batches may need a higher platform run timeout.

There are no automatic source-request or dataset-write retries. After an uncertain storage acknowledgement, the same run does not replay that output attempt. This favors avoiding duplicates over completeness: a reserved result may be missing after a crash. Check the Dataset and run report before investigating. A fresh run is a new audit; checking the same URL in a new run is a new snapshot.

JSON, CSV and automation

Use Apify Dataset export, or read the existing Dataset through the API:

GET https://api.apify.com/v2/datasets/DATASET_ID/items?format=json&clean=true
Authorization: Bearer YOUR_APIFY_TOKEN

Change format=json to format=csv for CSV. Nested lists and objects may expand into CSV columns, depending on export options. Import website-supplied text as text in spreadsheet applications; do not execute it as formulas. Keep API tokens in secrets, not in URLs or workflow logs.

For n8n, start an Actor run through the Apify integration, wait for completion, then read its Dataset and RUN_REPORT. Scheduling is optional and creates separate runs; no automatic schedule is configured by this release.

Pricing

Introductory price: $0.75 per 1,000 valid page snapshots ($0.00075 each), with no startup event. The result event applies only to valid snapshots saved to the default Dataset, not failed web requests or individual metadata fields. Check the Pricing tab for the currently effective settings.

An acknowledgement can be lost after a record was saved. That record may still be present and charged once; it is not automatically written or charged again by this Actor. Platform credits, free-plan rules and run limits still apply. Build/resource usage and event charges are different things.

Help with a run

Start with the Dataset and RUN_REPORT. ROBOTS_DENIED means the page was excluded by the site’s crawling rules; NOT_HTML means the response is not an HTML page; SOURCE_PROXY_UNSUPPORTED means the configured source proxy cannot provide the required connection safeguards. A spending-limit stop is not proof that a page lacks metadata. Include the run ID and affected URL when reporting a problem; keep tokens and private query parameters out of support messages.