LCP, INP, and CLS are the metrics Google grades. Here is the verdict, then the rubric.
Quick answer
DebugBear scores 90/100 for the job most teams actually have: continuously monitoring Core Web Vitals and getting alerted when LCP, INP, or CLS regress. It combines scheduled synthetic tests with real-user monitoring and starts around $49/month. If you want a free spot check, Google PageSpeed Insights is the runner-up and the best free tool — it shows Lighthouse lab data plus Google’s own CrUX field data. For the deepest single-page diagnosis, use WebPageTest.
The ranking
| Rank | Tool | Best for | Entry price | SR Score |
|---|---|---|---|---|
| 1 | DebugBear | Continuous CWV monitoring + alerts | From ~$49/mo | 90 |
| 2 | Google PageSpeed Insights | Free CWV spot checks (lab + CrUX) | Free | 88 |
| 3 | WebPageTest (Catchpoint) | Deep one-off diagnosis | Free; Pro from ~$15/mo | 85 |
| 4 | GTmetrix | Affordable scheduled monitoring | Free; PRO from ~$5–10/mo | 80 |
| 5 | Chrome DevTools (Lighthouse) | Free local debugging | Free | 78 |
How to choose
Start with what you are trying to learn. If you need to know whether real users are passing Core Web Vitals — the only thing Google’s page-experience signal cares about — you need field data (CrUX or your own RUM). PageSpeed Insights shows CrUX for free; DebugBear collects your own RUM continuously. If you need to know why a page is slow or unstable, you need lab data with diagnostics: WebPageTest’s filmstrips and waterfalls, or Lighthouse in DevTools. If you need to catch regressions before they reach users, you need monitoring with alerts — DebugBear or GTmetrix.
One trap worth naming: a green Lighthouse score in a lab does not mean your field metrics pass. Lab tests run on a controlled machine and connection; real users are on slower devices and networks, and INP in particular often looks fine in lab and fails in the field. Always confirm against CrUX before declaring victory.
Methodology
Weights: Field + lab evidence 30, Capability vs. claims 25, Value for money 20, Support & docs 15, Transparency 10.
Evidence is weighted highest and split deliberately: a CWV tool must surface both field data (CrUX/RUM, what Google uses) and lab data (Lighthouse/synthetic, reproducible for diagnosis). Capability rewards alerting, history, and INP-level detail. Value is measured at the plan a real team needs.
DebugBear — 90
Purpose-built for Core Web Vitals monitoring. It runs scheduled synthetic Lighthouse tests, collects real-user data (RUM), tracks LCP/INP/CLS over time, and alerts on regressions with request-level waterfalls to find the cause. Pricing starts around $49/month (entry tier with daily monitoring for a few sites), scaling to higher tiers for more pages and test frequency; the published range runs to roughly $899/mo at the top.
Trade-off: it is a paid product. For a single one-time check it is overkill versus free tools.
| Criterion | Score |
|---|---|
| Field + lab evidence (30) | 28 |
| Capability vs. claims (25) | 23 |
| Value for money (20) | 16 |
| Support & docs (15) | 14 |
| Transparency (10) | 9 |
Google PageSpeed Insights — 88
Free, no account, and authoritative. PSI runs Lighthouse (updated to Lighthouse 13.0 in October 2025) for lab data and pulls CrUX field data — the same 75th-percentile, 28-day field metrics Google references. It is the single best free way to see whether a URL passes Core Web Vitals.
Trade-off: it is a snapshot, not a monitor — no history, no alerts. Field data needs enough CrUX traffic to appear, so low-traffic URLs may show lab data only.
| Criterion | Score |
|---|---|
| Field + lab evidence (30) | 27 |
| Capability vs. claims (25) | 21 |
| Value for money (20) | 20 |
| Support & docs (15) | 12 |
| Transparency (10) | 8 |
WebPageTest (Catchpoint) — 85
The deepest single-test diagnostic tool: filmstrips, request waterfalls, multi-location and multi-device testing, and granular control. The Starter tier is free (300 test runs/month); Pro starts around $15/mo (annual) / $18.75 monthly for 1,000 runs, with an Expert plan that folds WPT into Catchpoint’s monitoring platform.
Trade-off: brilliant for diagnosis, less convenient for ongoing dashboards than DebugBear.
| Criterion | Score |
|---|---|
| Field + lab evidence (30) | 25 |
| Capability vs. claims (25) | 23 |
| Value for money (20) | 17 |
| Support & docs (15) | 12 |
| Transparency (10) | 8 |
GTmetrix — 80
A popular, affordable performance tester with scheduled monitoring slots. PRO plans start very low — Micro ~$5/mo, Solo ~$10/mo (annual), Starter ~$20/mo — adding monitored slots, hourly frequency, premium locations, and mobile testing.
Trade-off: GTmetrix’s headline scoring leans on Lighthouse lab data; for the CrUX field view that drives Google’s signal you still want PSI or RUM. Good value, narrower evidence.
| Criterion | Score |
|---|---|
| Field + lab evidence (30) | 21 |
| Capability vs. claims (25) | 20 |
| Value for money (20) | 18 |
| Support & docs (15) | 12 |
| Transparency (10) | 9 |
Chrome DevTools (Lighthouse) — 78
Free and built into Chrome. The Lighthouse panel runs lab audits; the Performance panel profiles INP and main-thread work locally. Indispensable for debugging on your own machine.
Trade-off: local lab only — no field data, no history, no alerting. It tells you why a page is slow, not whether your real users are affected.
| Criterion | Score |
|---|---|
| Field + lab evidence (30) | 20 |
| Capability vs. claims (25) | 20 |
| Value for money (20) | 20 |
| Support & docs (15) | 12 |
| Transparency (10) | 6 |
Re-weighting the rank
DebugBear takes #1 because we weight ongoing capability (monitoring, alerting, history) into the score and most teams need to keep CWV green, not check it once. If your job is a single audit and budget is zero, weight value to 40 and PageSpeed Insights wins outright — it is free and shows the field data Google uses. If your job is deep one-time diagnosis, weight diagnostic capability up and WebPageTest leads. None of these is wrong; they answer different questions, and the per-criterion tables let you recompute the rank for yours.
Verification
- DebugBear — pricing and CWV monitoring features verified on debugbear.com/pricing and /core-web-vitals-monitoring (from ~$49/mo; synthetic + RUM; LCP/INP/CLS).
- Google PageSpeed Insights — Lighthouse 13.0 update and CrUX field-data methodology verified on developers.google.com/speed/docs/insights and release notes (free; 75th percentile / 28-day field window).
- WebPageTest — free Starter (300 runs) and Pro from ~$15/mo annual verified on catchpoint.com/pricing and TrustRadius pricing listing.
- GTmetrix — PRO tiers (Micro ~$5, Solo ~$10, Starter ~$20) verified on gtmetrix.com/pricing.
- Chrome DevTools / Lighthouse — free, in-browser; Lighthouse documentation (developers.google.com).
Related rankings
- Best Page Speed Testing Tools (2026): Scored & Ranked
- Best Redirect Checker Tools (2026): .htaccess & 301 Ranked
- Best Image Optimization Tools (2026): Scored & Ranked
- Best Indexing Tools for SEO (2026): Scored & Ranked
Frequently asked questions
- What are the 2026 Core Web Vitals?
- Largest Contentful Paint (LCP, loading), Interaction to Next Paint (INP, responsiveness — which replaced FID in March 2024), and Cumulative Layout Shift (CLS, visual stability). Google's field tools report them at the 75th percentile over a rolling 28-day window via CrUX.
- What is the best Core Web Vitals tool?
- For continuous monitoring, DebugBear (90/100) — it tracks LCP, INP, and CLS with synthetic and real-user data and alerts on regressions, from about $49/month. For a free check, Google PageSpeed Insights combines Lighthouse lab data with CrUX field data.
- What is the best free Core Web Vitals tool?
- Google PageSpeed Insights — free, no account, and the only tool here that shows both Lighthouse lab scores and Google's own CrUX field data, which is the data Google uses for the page-experience signal.
- Lab data vs. field data — which matters?
- Field data (CrUX/RUM) is what Google uses for ranking; lab data (Lighthouse/WebPageTest) is reproducible and best for diagnosing causes. Good workflows use both: field to know if you have a problem, lab to find why.