Core Web Vitals Monitor: bulk PageSpeed & CrUX checker (GTmetrix alternative)

Get alerted when your speed app or plugin stops working.

Audit hundreds of URLs with PageSpeed Insights and CrUX real-user data (with a free Google API key) in one run, track Core Web Vitals (LCP, INP, CLS), get Slack alerts on regressions and benchmark competitors. It flags pages that score 90+ in Lighthouse while real Chrome users fail Core Web Vitals. This can mean a speed plugin or app is optimising for the PageSpeed test rather than for real visitors, or that the problem only shows on real devices.

A GTmetrix alternative with no monthly plan: $0.008 per URL per device, competitor URLs included, plus $0.01 per own URL × device benchmarked. Google's PSI quota still applies (25,000 requests/day with a free key).

Use cases

• SEO consultants: weekly client reports with regression alerts.
• Shops using a speed app (NitroPack, WP Rocket): daily checks for real users.
• E-commerce teams: benchmark key templates against competitors.
• Developers: check every release from CI, n8n, Make or Zapier.

Sample output (shortened, real coolblue.nl run):

{
  "url": "https://www.coolblue.nl/",
  "strategy": "mobile",
  "performanceScore": 61,
  "field": { "source": "crux-api", "lcpMs": 2002, "inpMs": 202, "cls": 0.01 },
  "coreWebVitalsAssessment": "failed",
  "labFieldMismatch": false,
  "failingFieldMetrics": [{ "metric": "INP", "p75": 202, "rating": "needs-improvement" }]
}

What does it do?

• Bulk PageSpeed Insights (v5). Mobile, desktop or both, with retries and rate-limit handling.
• CrUX field data. Real-user p75 LCP, INP, CLS, FCP and TTFB from the official CrUX API. Falls back to origin-level data when a URL has too little traffic; field.scope says url or origin. Optionally adds 6-month CrUX history.
• Core Web Vitals assessment (passed / failed / insufficient data) on field data.
• Threshold checks for LCP, INP, CLS and the performance score, using your thresholds. Field data is preferred; each check names its source.
• History and regression detection. Each run is compared with the previous one for the same URL and device, with deltas and a regressions / improvements list (for example "Performance score dropped 72 → 60" or "Lab LCP worse by 38% (2100ms → 2900ms)").
• Lab-vs-field mismatch warning. When the Lighthouse score is 90 or higher but CrUX fails Core Web Vitals (p75 LCP > 2.5 s, INP > 200 ms or CLS > 0.1), the row gets labFieldMismatch: true, the failing metrics in failingFieldMetrics and a plain-language message in warnings. They also appear in the report, OUTPUT and alerts.
• Competitor comparison. Your rank, the gap to the best competitor's score and LCP, and a side-by-side table in the report.
• Alerts to any webhook or Slack Incoming Webhook.
• Works without an API key, with limits (see FAQ).

Input

Field	Description	Default
urls	URLs to audit (yours)	–
competitorUrls	Competitor URLs to benchmark against	–
strategy	mobile, desktop or both	mobile
pageSpeedApiKey	Your Google API key (stored as a secret). Recommended.	–
includeCrux / includeCruxHistory	Real-user data from the CrUX API (needs an API key)	true / false
engine	auto (PSI, then local Lighthouse when the quota is exhausted), pagespeed, local	auto
alertThresholds	{ lcpMs, inpMs, cls, performanceScore }	2500 / 200 / 0.1 / 90
historyMode, historyStoreName	Keep history and compute deltas and regressions	true
regressionScoreDrop, regressionMetricPct	What counts as a regression	5 points / 15%
webhookUrl, alertOn	Where to send alerts and for what (regressions, threshold-failures, both)	– / regressions
maxConcurrency, locale	Advanced settings	auto / en

Free Google API key (2 minutes): in Google Cloud Console, create a project, enable PageSpeed Insights API and Chrome UX Report API, create an API key under Credentials restricted to those two APIs, and paste it into pageSpeedApiKey. The default PSI quota is 25,000 requests per day.

Output

One dataset row per URL × device, also with role (own or competitor), status (ok or error), full lab/field metrics, checks, passedAllThresholds, topOpportunities, engine and pageSpeedInsightsUrl. On the next run of the same URL, previous, delta, regressions and improvements are filled in.

labFieldMismatch is false in the sample because the lab score (61) is below 90. When a page scores 90+ but field data fails, it is true and warnings explains why, for example: "Lab score 98 looks good, but real Chrome users fail Core Web Vitals: CLS 0.11 (needs improvement)..." It is null when there is no field data or no lab score.

The key-value store holds report.html, OUTPUT (run summary, labFieldMismatches, alerts) and COMPARISON. Download the dataset as CSV, Excel or JSON.

Schedule recurring Core Web Vitals monitoring

Speed usually slips after launch, so monitoring is where this Actor is most useful:

1. Save your URLs, competitors, API key and webhookUrl as an Apify Task (one per client or site, each with its own historyStoreName).
2. In Apify Console, open Schedules and add the task with a cron expression, for example 0 7 * * * for daily at 07:00 or 0 7 * * 1 for Mondays.
3. alertOn defaults to regressions, so Slack only hears from it when something got worse. A lab-vs-field mismatch is alerted only when it is new since the previous run.

Use an API key so the engine, and the numbers, stay the same between runs. See Apify schedules.

Use it from n8n, Make and Zapier

Apify has official integrations for all three: connect your Apify account with an API token, then use the Apify app or node to run an Actor or Task and fetch its dataset items, or trigger a workflow when a run finishes.

Typical workflows:

• Client dashboard: after each scheduled run, keep rows with role: "own" and status: "ok" and append url, strategy, performanceScore, field LCP/INP/CLS and coreWebVitalsAssessment to a Google Sheet or Looker Studio source.
• Ticket on regression: for role: "own" rows with a non-empty regressions list, create a Jira, Linear or Trello card. For mismatches, use the webhook alerts (new mismatches only, by default), since labFieldMismatch stays true on every run until the page is fixed.
• After a deploy: run the Task for the changed templates and post the result to the release channel.

Guides: n8n, Make, Zapier.

Pricing (pay per event)

No subscription, no auto-renew, no monthly plan, no scan quota or page-view tiers. You pay only for audits that run.

Event	Price
URL audited (url-audited), per URL × device	$0.008
Competitor comparison (competitor-comparison), per own URL × device benchmarked	$0.01

Competitor URLs are audited too, so each one costs $0.008 per device, on top of the $0.01 comparison per own URL × device.

Examples: a weekly check of 20 URLs on mobile costs about $0.64 a month. 100 URLs on mobile and desktop, daily, cost about $48 a month. Failed audits are not charged.

Apify's free monthly credit covers hundreds of audits. Set Maximum cost per run to cap spend.

FAQ

Do I need a Google API key? No, but it is recommended, and needed for heavy use. Without a key, Google's keyless quota is shared by everyone and is often exhausted. The Actor then uses its built-in Lighthouse 12 fallback, which gives lab data only and no CrUX field data. With your own free key, results come from Google's PageSpeed Insights API itself, plus CrUX real-user data.

Why do scores differ from my last run? Lighthouse lab scores vary by about ±5 points between runs, so the default regression thresholds ignore small changes and field data (CrUX, 28-day p75) is preferred for pass/fail.

Why is INP empty? INP can only be measured on real users, so it comes only from CrUX. Low-traffic pages have no CrUX data; the origin-level value is used when available.

My speed plugin gives me 95+ in PageSpeed. Why does this flag a mismatch? The PageSpeed score is one lab test on a simulated device. Core Web Vitals pass/fail comes from real Chrome users over 28 days. Some plugins improve the lab test (for example by delaying scripts) without helping visitors. When the two disagree, trust the field data and test on a real phone. Field data lags by up to 28 days, so give a change a few weeks.

How is this different from GTmetrix? It is pay per use with no monthly plan, audits hundreds of URLs per run, and adds CrUX data, regression alerts and competitor benchmarks. It does not offer GTmetrix's choice of test locations or its waterfall and video views.

Can I send alerts to Slack? Yes. Paste a Slack Incoming Webhook URL into webhookUrl. The payload includes a Slack text field and structured alerts.

Limitations

• The built-in Lighthouse fallback runs on Apify's servers. Mobile scores can be lower than PSI's (different CPU).
• Sites with aggressive bot protection may block the fallback. PSI with an API key is usually not blocked.
• CrUX data exists only for URLs and origins with enough Chrome traffic.
• The PSI API has a limit of about 60 s per audit. Very slow pages can time out and are reported as failed, not charged.

Keywords

Core Web Vitals, Core Web Vitals monitor, PageSpeed checker, bulk PageSpeed Insights, CrUX, Chrome UX Report, LCP INP CLS, site speed monitoring, GTmetrix alternative, competitor speed comparison.