Lighthouse Bulk Audit: Scores, Core Web Vitals, Top Fixes

Lighthouse bulk audit: scores, Core Web Vitals and top fixes for a list of URLs

Paste a list of pages. For each one, the Actor runs Google Lighthouse in headless Chrome and returns one row with:

• Four scores from 0 to 100: Performance, Accessibility, Best Practices and SEO.
• Lab metrics: Largest Contentful Paint (LCP), Cumulative Layout Shift (CLS), Total Blocking Time (TBT), First Contentful Paint (FCP), Speed Index and Time to First Byte (TTFB).
• Top fixes: the changes Lighthouse says matter most, with the time each speed fix saves.

$0.01 per URL. Pages that do not load, that Lighthouse cannot score, or that robots.txt blocks are free.

Quick start

1. Paste URLs or domains into Pages to audit, one per line.
2. Pick Mobile, Desktop or Both.
3. Click Start. Each page takes about 10 to 30 seconds, one page at a time.
4. Download the table as CSV, Excel or JSON, or read it through the API.

What can you do with it?

• Check a whole site after a release: paste your key pages and compare the scores with last week's run.
• Find the slowest pages: sort by performance or lcpMs and fix those first.
• Audit client or prospect sites: an agency can show a list of sites with their scores and the top 3 fixes for each.
• Schedule it: run it every week with Apify Schedules and keep a history of your scores.
• AI agents: call the Actor as a tool (Apify API or Apify MCP server) to answer "Which of these 40 landing pages has the worst LCP on mobile?".

Input example

{
    "urls": ["https://www.spacejam.com/1996/", "https://example.com"],
    "device": "mobile",
    "topFixes": 3
}

Field	What it does
urls	Required. Domains or http(s) URLs. A domain is tested over HTTPS. Duplicates are removed. Other schemes are ignored.
device	mobile (default), desktop, or both. both makes 2 rows and 2 charges per URL.
topFixes	How many fixes to list per page, 0 to 10. Default 3.

Output example

One item per URL and device (from a real run):

{
    "url": "https://www.spacejam.com/1996/",
    "finalUrl": "https://www.spacejam.com/1996/",
    "device": "mobile",
    "status": "ok",
    "performance": 99,
    "accessibility": 85,
    "bestPractices": 88,
    "seo": 91,
    "lcpMs": 2003,
    "cls": 0,
    "tbtMs": 3,
    "fcpMs": 945,
    "speedIndexMs": 1043,
    "ttfbMs": 600,
    "topFixes": [
        { "id": "modern-http-insight", "category": "performance", "title": "Modern HTTP", "savingsMs": 650 },
        { "id": "cache-insight", "category": "performance", "title": "Use efficient cache lifetimes", "savingsMs": 200 },
        {
            "id": "html-has-lang",
            "category": "accessibility",
            "title": "`<html>` element does not have a `[lang]` attribute",
            "savingsMs": null
        }
    ],
    "lighthouseVersion": "13.5.0",
    "benchmarkIndex": 3886,
    "error": null,
    "auditedAt": "2026-10-21T08:30:00.000Z"
}

Field	Meaning
status	ok (scored and charged), skipped (robots.txt, free) or error (did not load or could not be scored, free).
performance, accessibility, bestPractices, seo	Lighthouse category scores, 0 to 100.
lcpMs, tbtMs, fcpMs, speedIndexMs, ttfbMs	Lab metrics in milliseconds. These are the values Lighthouse scores.
cls	Cumulative Layout Shift (no unit).
topFixes	Speed fixes come first, largest time saved first. savingsMs is Lighthouse's estimate of the time saved on LCP, FCP or TBT, whichever is largest. Then failed checks from Accessibility, Best Practices and SEO, the ones that cost the most points first.
finalUrl	The page Lighthouse ended on, after redirects.
benchmarkIndex	Lighthouse's measure of the test machine's CPU speed. Useful when you compare runs.
error, skipReason	Why a page has no scores.

Lab data, not field data

Lighthouse measures one page load in a test browser. That is lab data. It is good for finding problems and for comparing runs over time.

• INP is not measured. Interaction to Next Paint needs real people clicking. TBT is the lab metric that tracks it best.
• Scores move between runs. A few points of difference between two runs of the same page is normal. Compare trends, not single numbers.
• PageSpeed Insights can show different numbers. It runs on Google's machines, and it also shows field data from real Chrome users (the Chrome UX Report). This Actor shows lab data only.

Mobile runs use Lighthouse's default mobile settings: a mid-range phone on a slow 4G connection, simulated. Desktop runs use Lighthouse's desktop settings.

Pricing

• $0.01 per URL audited (per URL and device, so both costs $0.02 per URL).
• Pages that robots.txt blocks, that do not load, or that Lighthouse cannot score are not charged.
• A tiny Actor start event (Apify's standard start charge) applies once per run.
• Set a maximum total charge on the run and the Actor stops before it goes over.

Apify free plan: each run audits the first 10 URLs. Upgrade to a paid Apify plan to audit any number in one run.

Polite and safe

• The Actor reads each site's robots.txt first. A page that robots.txt blocks for Lighthouse (Chrome-Lighthouse) or for all bots (*) is skipped.
• It audits one page at a time, so a site never gets more than one Lighthouse visit at once.
• It loads public internet pages only. Every request the browser makes (the page, each redirect, each image or script) goes through a check that refuses private and local network addresses.

FAQ

How long does a run take? About 10 to 30 seconds per page, one page at a time. Heavy pages take longer. A page that takes over 2 minutes is stopped and not charged.

Can I audit pages behind a login? No. The Actor tests public pages only, with no cookies.

Which Lighthouse version? The version used is in lighthouseVersion on every row.

Does it use PageSpeed Insights? No. It runs Lighthouse itself, so there is no Google API key and no PageSpeed Insights quota.