On-Page SEO Audit - Metadata, Headings and Explained Scores

On-page SEO scraper returns page metadata and a rule-based audit for supplied public URLs. On-page SEO API results include titles, descriptions, headings, links, indexability signals, scores, and errors.

Input example

{
  "urls": [
    "https://example.com/"
  ],
  "maxItems": 3
}

The input keys and allowed values follow .actor/input_schema.json.

Output fields

Field	Type	Example or excerpt from sample
url	string	"https://example.com/"
finalUrl	string	"https://example.com/"
statusCode	integer	200
responseTimeMs	integer	297
fetchedAt	string	"2026-10-05T13:05:03.966343+00:00"
title	string	"Example Domain"
titleLength	integer	14
titleCount	integer	1
metaDescription	string or null	null
metaDescriptionLength	integer	0
metaDescriptionCount	integer	0
canonical	string or null	"https://schema.org/Product"
canonicals	array	[]
robotsMeta	object	{}
xRobotsTag	string or null	null
noindex	boolean	false
hreflang	array	[]
headings	array	[]
h1Count	integer	0
imageCount	integer	0
imagesWithAlt	integer	0
imagesWithEmptyAlt	integer	0
imagesMissingAlt	integer	0
imageAltCoverage	integer or null	100
internalLinkCount	integer	0
externalLinkCount	integer	0
structuredDataTypes	array	[]
structuredDataErrors	array	[]
openGraph	object	{}
twitter	object	{}
wordCount	integer	25
error	string or null	null
scoreVersion	string	"1.0"
scoreRules	array	[{"id": "title", "maximum": 15, "earned": 0, "passed": false, "reason": "Exactly one …
score	integer	20
issues	array	[]
attempts	integer	1
redirectChain	array	[]

Output example

Trimmed from sample_output.json; omitted fields remain available in the dataset.

{
  "url": "https://example.com/",
  "finalUrl": "https://example.com/",
  "statusCode": 200,
  "responseTimeMs": 297,
  "fetchedAt": "2026-10-05T13:05:03.966343+00:00",
  "title": "Example Domain",
  "titleLength": 14,
  "error": null
}

Transparent scoring: version 1.0

Rule	Maximum	Full-credit condition
Title	15	Exactly one title, 30–60 characters
Description	15	Exactly one description, 70–160 characters
Canonical	10	Exactly one canonical resolving to an absolute HTTP(S) URL
Indexability	10	No noindex/none detected in the inspected meta tags or X-Robots-Tag
H1	10	Exactly one non-empty H1
Heading order	5	Starts with H1; adjacent headings do not skip upward by more than one level
Image alt	10	Non-empty alt on all images; partial points = 10 × non-empty-alt ratio; no images earns 10
Text quantity	10	At least 300 text tokens
Structured data	5	At least one JSON-LD/microdata type, without JSON-LD parse errors
Open Graph	5	Non-empty og:title, og:description and og:image
Twitter card	5	twitter:card is summary, summary_large_image, app or player

The maximum sums to 100. Image partial credit is rounded to two decimals; the other rules award their maximum or zero. scoreRules contains the point calculation and reason for each rule.

These thresholds are this Actor's checklist heuristics, not Google's ranking rules or a ranking forecast. Empty alt can be correct for a decorative image; it is reported separately and receives no non-empty-alt credit. Short pages or pages intentionally marked noindex can be correct for their purpose. Inspect the reasons before choosing changes. Response time and hreflang presence do not contribute points.

Limits and error handling

Input	Default	Allowed range
maxItems	1,000	1–10,000 result rows, including errors
concurrency	5	1–20 workers
timeoutSecs	30	5–120 seconds per fetch attempt
maxRetries	2	0–3 retries after the initial attempt
maxResponseBytes	2,000,000	1,024–10,000,000 decoded bytes per page

The input list may contain at most 100,000 URL entries before deduplication. The first distinct URLs within maxItems are attempted; dataset order follows completion order. Lower concurrency when requesting many pages from one host.

Network errors and HTTP 408/429/500/502/503/504 receive exponential backoff (1, 2, 4 seconds). Numeric Retry-After is honored up to 30 seconds. There is a maximum of ten redirects per attempt. Redirect destinations are validated before requesting them. Unsupported content types and excessive responses produce error rows.

Only explicitly supplied pages and their HTTP redirects are requested. There is no site-wide crawl, browser rendering, login, CAPTCHA solving, robots.txt fetch, hreflang reciprocity check, link-status check or schema semantic validation. Public host validation rejects private IP literals and private DNS answers; DNS preflight is not connection-level IP pinning, so this is not a security gateway for untrusted hostile inputs. HTML parsing is best-effort for malformed markup, and CSS visibility is not evaluated.

FAQ

What are the output limits? maxItems caps result rows, including errors. Fetch failures, non-HTML responses, and oversized pages produce error rows.

Do I need a proxy or login? Public URLs need no login or proxy. The actor does not authenticate or render JavaScript.

How is this priced? Proposed pay per event: $0.002 per successfully audited page row. Error rows are uncharged. These are proposals in PRICING.md; the owner must configure Store billing.

How are rate limits handled? concurrency limits fetches; transient failures are retried with backoff and Retry-After.

How fresh is the data? Results reflect HTML fetched at run time. JavaScript rendered content may be absent. Scores are this actor’s checklist heuristics, not search ranking predictions; review rule reasons before acting.

Use with AI agents / MCP

After the actor is published, connect your agent to the Apify MCP server with an Apify API token. Use call-actor with the published actor identifier and the JSON input shown above. Read the returned dataset with get-dataset-items; the run response itself contains status and storage IDs, not the result rows. Alternatively, use the Apify API to start a run and read its default dataset. Keep the actor input bounded with maxItems, and inspect source links and any error fields before using results in an answer.