GuGuData.io Guides

Real-Browser PageSpeed and SEO Audit API Guide

Run mobile or desktop laboratory audits for public webpages and receive page quality scores, loading metrics, audit findings, and prioritized opportunities.

5 min read
Try the live demoView API details →

Real-Browser PageSpeed and SEO Audit API Guide

The Technical SEO Diagnostics API loads a public webpage in a browser-based laboratory environment and returns requested performance, accessibility, best-practice, and SEO scores. It also provides loading measurements, individual audit results, and prioritized opportunities that teams can add to release checks, monitoring dashboards, and technical SEO workflows.

Try the live demo or review the product details.

API overview

Item Value
Method GET
Endpoint https://api.gugudata.io/v1/websitetools/pagespeed-score
Device profiles desktop, mobile
Score categories Performance, accessibility, best practices, SEO
Result cache 15 minutes unless forceRefresh=true

Request parameters

Parameter Type Required Default Description
appkey string Yes None GuGuData application key.
url string Yes None Complete public HTTP or HTTPS URL, up to 2,048 characters.
strategy string No desktop Browser profile: desktop or mobile.
locale string No en Locale used for audit labels.
categories string No All four Comma-separated subset of performance, accessibility, best-practices, and seo.
forceRefresh boolean No false Replace a recent cached result with a fresh audit.

Always URL-encode the complete value of url:

curl -G "https://api.gugudata.io/v1/websitetools/pagespeed-score" \
  --data-urlencode "appkey=YOUR_APPKEY" \
  --data-urlencode "url=https://gugudata.github.io/gugudata-io/guides/pagespeed-and-seo-score-api-guide/" \
  --data-urlencode "strategy=mobile" \
  --data-urlencode "locale=en" \
  --data-urlencode "categories=performance,accessibility,best-practices,seo"

Response structure

The endpoint keeps the same top-level data fields for compatibility. url is the final audited URL after redirects. Scores are integers from 0 to 100, and only requested categories appear in scores.

{
  "dataStatus": {
    "requestParameter": "url=https://gugudata.github.io/gugudata-io/guides/pagespeed-and-seo-score-api-guide/&strategy=mobile&locale=en&categories=performance,accessibility,best-practices,seo&forceRefresh=false",
    "statusCode": 200,
    "status": "SUCCESS",
    "statusDescription": "successfully",
    "responseDateTime": "2026-09-04 09:00:00.000",
    "dataTotalCount": 1
  },
  "data": {
    "url": "https://gugudata.github.io/gugudata-io/guides/pagespeed-and-seo-score-api-guide/",
    "strategy": "mobile",
    "locale": "en",
    "checkedAt": "2026-09-04T01:00:00.000Z",
    "coreMetrics": {
      "statusCode": 200,
      "fetchTimeMs": 180,
      "contentBytes": 48216,
      "contentType": "text/html",
      "compressed": true,
      "imageCount": 2,
      "missingImageAlt": 0,
      "scriptCount": 4,
      "url": "https://gugudata.github.io/gugudata-io/guides/pagespeed-and-seo-score-api-guide/",
      "firstContentfulPaintMs": 940,
      "largestContentfulPaintMs": 1210,
      "totalBlockingTimeMs": 15,
      "cumulativeLayoutShift": 0.01,
      "speedIndexMs": 1080,
      "interactiveMs": 1280
    },
    "scores": {
      "performance": 96,
      "accessibility": 100,
      "bestPractices": 96,
      "seo": 100
    },
    "audits": [
      {
        "id": "first-contentful-paint",
        "title": "First Contentful Paint",
        "passed": true,
        "value": "0.9 s"
      }
    ],
    "opportunities": []
  }
}

The numbers above illustrate field shapes only. Browser laboratory results can vary with page content, network conditions, and third-party resources, so automated checks should use ranges and trends instead of exact score equality.

Loading measurements

Field Meaning
firstContentfulPaintMs Time until the first page content is painted.
largestContentfulPaintMs Time until the largest measured content element is painted.
totalBlockingTimeMs Time that long main-thread tasks block interaction.
cumulativeLayoutShift Laboratory layout stability measurement.
speedIndexMs How quickly visible content appears during the load.
interactiveMs Laboratory estimate for page interactivity.

These are laboratory measurements, not real-user field data. They do not include a real-user Interaction to Next Paint measurement and do not guarantee search ranking changes.

Practical workflows

Release regression checks

Audit a stable set of representative URLs before and after a deployment. Store the URL, strategy, requested categories, checked time, scores, metrics, and failed audits. Compare like-for-like profiles and investigate meaningful changes instead of failing a build on a one-point score fluctuation.

Mobile and desktop monitoring

Run separate mobile and desktop checks when templates or resource loading differ by device. Never combine the two strategies in one time series without preserving the strategy field.

Prioritized technical SEO work

Join audit findings with search impressions, conversions, or page importance. High-value pages with repeated failed audits should usually be reviewed before low-traffic pages with the same issue.

Fresh checks after deployment

Routine dashboards should reuse the default cache. Use forceRefresh=true after a deployment or an important page change when a new browser run is required immediately.

HTTP status handling

HTTP status Meaning Recommended action
200 Browser audit completed. Store the normalized result.
400 URL or request option is invalid. Correct the request before retrying.
401 Application key is missing or unknown. Check appkey.
403 The key cannot access this product. Check authorization or subscription.
422 The target could not produce a usable audit. Confirm that it serves an auditable HTML page.
429 API rate limit reached. Reduce request concurrency.
502 The target failed during loading or auditing. Check the target and retry later if the failure is temporary.
503 Audit capacity or service is temporarily unavailable. Retry with bounded backoff.

Integration guidance

  • Keep appkey on your server rather than in browser code.
  • Queue large URL sets and respect rate limits.
  • Persist the final returned url, because redirects may change the audited page.
  • Treat nullable loading measurements as unavailable evidence, not as zero.
  • Save audit IDs and opportunities so teams can see why a score changed.
  • Use manual review for high-impact decisions; automated scores are diagnostic evidence rather than business outcomes.