Lighthouse Audit – Bulk PageSpeed & Core Web Vitals, No API Key

Run Google Lighthouse on any list of URLs and get performance, accessibility, best practices and SEO scores, Core Web Vitals (LCP, CLS, TBT, FCP, Speed Index, TTI) and the top fixes for every page, on mobile and desktop. It runs the open-source Lighthouse 13 engine in a real headless Chrome inside the Actor, so you don't need a PageSpeed Insights API key and Google's API quota doesn't apply. Paste URLs, click Start, and get one clean row per page. You can also save the full Lighthouse HTML report.

What it does

• Bulk Lighthouse audits: one URL or thousands. Each audit runs in its own fresh Chrome (like the Lighthouse CLI), with a hard timeout and automatic retry when Chrome crashes.
• No API key, no quota: Lighthouse runs locally, so nothing goes through the PageSpeed Insights API.
• Mobile, desktop or both: mobile is a Moto G Power with slow 4G (what PageSpeed Insights shows first). Desktop is 1350×940 on a fast connection. both gives two rows per URL.
• Scores 0–100 for performance, accessibility, best practices and SEO. Pick only the categories you need.
• Core Web Vitals and lab metrics: LCP, FCP, CLS, TBT, Speed Index and TTI, as numbers in ms.
• Top 5 opportunities from Lighthouse's insights and diagnostics, sorted by estimated savings: ms of LCP/FCP, ms of TBT, and bytes. Examples: "Reduce unused JavaScript – 908 KiB", "Render-blocking requests – 1,350 ms".
• Failed audits: a count per category, plus the titles of failing accessibility, best-practice and SEO checks (for example "Document does not have a meta description").
• Full reports (optional): the interactive Lighthouse HTML report and/or the complete JSON result, saved with a public link.
• Throttling: simulated (default, the same method PageSpeed Insights and the Lighthouse CLI use), devtools (applied throttling) or none.
• Errors are free: unresolvable domains, HTTP error pages, hung pages and timeouts become error rows that you are not charged for.

Use cases

• SEO and agency reporting: audit all key pages of a client site every month and track the scores in a spreadsheet or dashboard.
• Core Web Vitals monitoring: schedule a daily run on your top landing pages and get alerted when LCP or CLS gets worse after a deploy.
• Competitor benchmarking: compare the performance, SEO and accessibility of 100 competitor homepages in one run.
• Lead generation for web agencies: find slow or inaccessible websites in a list of prospects and attach the HTML report to your pitch.
• Accessibility checks: get the failing axe-core checks (contrast, labels, alt text) for many pages at once.
• CI / pre-release checks: call the API after each staging deploy and fail the build when a score drops below your budget.

Input example

{
  "urls": ["https://apify.com", "https://example.com", "en.wikipedia.org/wiki/Web_performance"],
  "formFactor": "both",
  "categories": ["performance", "accessibility", "best-practices", "seo"],
  "throttling": "simulated",
  "saveHtmlReport": true,
  "saveJsonReport": false,
  "timeoutSecs": 120
}

Other options: maxConcurrency (1–3 parallel audits, default 1 for the most accurate scores), maxRetries (default 1, only for Chrome crashes and blank pages), and maxUrls (a cap for big lists). Bare domains get https:// added, and duplicates are removed.

Output example

{
  "url": "https://github.com/",
  "finalUrl": "https://github.com/",
  "formFactor": "mobile",
  "scores": { "performance": 54, "accessibility": 100, "bestPractices": 96, "seo": 100 },
  "metrics": { "lcpMs": 11216, "fcpMs": 9866, "cls": 0.072, "tbtMs": 129, "speedIndexMs": 9866, "ttiMs": 16416 },
  "opportunities": [
    { "id": "unused-javascript", "title": "Reduce unused JavaScript", "displayValue": "Est savings of 908 KiB",
      "estimatedSavingsMs": 2400, "estimatedSavingsBytes": 930198, "tbtSavingsMs": 0, "score": 0 },
    { "id": "render-blocking-insight", "title": "Render-blocking requests", "displayValue": "Est savings of 1,350 ms",
      "estimatedSavingsMs": 1350, "estimatedSavingsBytes": 0, "tbtSavingsMs": 0, "score": 0 }
  ],
  "failedAudits": { "performance": 10, "accessibility": 0, "bestPractices": 1, "seo": 0 },
  "failedAccessibilityAudits": [],
  "failedBestPracticesAudits": ["Issues were logged in the `Issues` panel in Chrome Devtools"],
  "failedSeoAudits": [],
  "throttling": "simulated",
  "lighthouseVersion": "13.5.0",
  "chromeVersion": "154.0.0.0",
  "benchmarkIndex": 3856,
  "runWarnings": [],
  "fetchedAt": "2026-10-11T11:51:26.010Z",
  "reportUrl": "https://api.apify.com/v2/key-value-stores/<storeId>/records/00004-github-com-mobile.html",
  "jsonReportUrl": "https://api.apify.com/v2/key-value-stores/<storeId>/records/00004-github-com-mobile.json",
  "durationMs": 17536,
  "status": "ok",
  "error": null
}

The Output tab has three views: Scores & Core Web Vitals (one table row per URL and device), Opportunities & failed audits, and Errors. OUTPUT in the key-value store holds a run summary. See SAMPLE_OUTPUT.json for full rows, including error rows.

Pricing

Pay per event: $0.02 per completed audit (one URL on one device), plus $0.00005 per run start. That is $20 per 1,000 audits. The price is the same whether you run 1 or 4 categories, and with or without the HTML/JSON reports. Failed audits, invalid URLs and retries are free. The Actor respects your Maximum cost per run: it never starts an audit it could not pay for, and it stops cleanly when the limit is reached.

Actor	Price	How it works
Lighthouse Audit (this)	$0.02 / audit, $0.00005 / start	Lighthouse 13 in local Chrome, no API key. Mobile + desktop, 4 categories, CWV metrics, top-5 opportunities, failed-audit lists, HTML + JSON reports, errors free
dev00/Google-PageSpeed-Insights-api	$0.002 / query	Uses the PageSpeed Insights API
constant_quadr/lighthouse-auditor	Free Actor, you pay platform usage
perryay/website-performance-auditor	$0.025 / audit
nexgendata/page-speed-analyzer	$0.10 / audit
alizarin_refri/google-lighthouse-checker	$0.08 / audit

Competitor prices from the Apify Store, October 2026. Check their pages for current terms.

FAQ

Why are my scores different from PageSpeed Insights? PageSpeed Insights runs the same Lighthouse engine, but on Google's machines and from Google's network location. Lab scores always vary a few points between runs and machines, especially for performance. This Actor uses the same simulated throttling and device settings as PSI, so the results are directly comparable to each other. For trend tracking, always use the same settings and memory. PSI's "field data" (CrUX, real users) isn't included, because that comes from Google's API.

Which memory should I use? The default 4096 MB gives Lighthouse one full CPU core, which gives the most stable performance scores. 2048 MB works and is a bit cheaper on compute, but audits are slower and performance scores are noisier. Use 8192 MB with maxConcurrency: 2 to audit big lists faster.

How long does one audit take? About 8–15 seconds for a typical page and 20–35 seconds for very heavy news or e-commerce homepages. A list of 100 URLs on mobile takes about 20–30 minutes at concurrency 1.

Some sites return an error or a 403 page. A few sites block headless browsers or datacenter IPs. If Chrome only gets an error page, you get a free error row (for example CHROME_INTERSTITIAL_ERROR or ERRORED_DOCUMENT_REQUEST). If the site serves a bot-check page with status 200, Lighthouse audits that page.

Can I audit pages behind a login or on localhost? No. Only publicly reachable URLs are supported.

Where are the HTML reports and how long are they kept? In the run's default key-value store. Every row has a direct reportUrl that opens the full interactive Lighthouse report in the browser. Retention follows your Apify plan's storage settings. Turn on saveJsonReport if you need every audit's raw details (for example to load them into the Lighthouse Viewer).

Is it legal? Lighthouse loads a page once, like a single browser visit. Only audit sites you are allowed to test, and keep the list reasonable for small sites.

Use with AI agents / Apify MCP

• Apify MCP server: add the Actor to Claude Desktop, Cursor or VS Code with https://mcp.apify.com?actors=gazidev/lighthouse-audit. An agent can call it with {"urls":["https://mysite.com"],"formFactor":"both"} and get back scores, Core Web Vitals and the top fixes as compact JSON, which is small enough to reason over, unlike the raw Lighthouse report.
• API (synchronous): POST https://api.apify.com/v2/acts/gazidev~lighthouse-audit/run-sync-get-dataset-items?token=... with the input JSON returns the rows in one call (about 10–20 s per URL).
• Scheduled monitoring: save your URLs as a Task, schedule it daily or weekly, and send the dataset to Google Sheets, Slack or a webhook with Apify integrations.

More from the website toolkit

• SEO Audit Crawler: broken links, meta tags, redirects and on-page SEO issues across a whole site.
• Website Screenshot API: bulk full-page screenshots and PDFs.
• Tech Stack Detector: find out which CMS, frameworks and tools a website uses.

Lighthouse is an open-source project by Google (Apache-2.0). This Actor isn't affiliated with or endorsed by Google.