The best free website speed tests, ranked
By Philipp Kant 7 min read
A website speed test is a tool that loads a page, measures how fast it becomes visible and usable, and reports metrics such as Largest Contentful Paint along with what to fix. Free speed tests fall into two groups: lab tools that load the page in a controlled environment, and field tools that report what real visitors actually experienced. The difference decides what the number means.
We ranked the well-known free options by how much of the real question each one answers: what is slow, why, and what to fix first. The score a tool prints matters less than the diagnosis behind it.
| Tool | Type | Data source | Free tier |
|---|---|---|---|
| PageSpeed Insights | Lab + field | Lighthouse + Chrome UX Report | Fully free, no signup |
| WebPageTest | Lab | Real browsers, real locations | Free account, limited runs |
| Lighthouse (DevTools) | Lab | Your own machine | Fully free, built into Chrome |
| GTmetrix | Lab | Lighthouse-based | Free account, one location |
| Pingdom | Lab | Own grading model | Fully free test page |
1. PageSpeed Insights, the reference point
Google PageSpeed Insights is where to start: paste a URL and get a Lighthouse lab run from Google’s infrastructure plus, when the site has enough traffic, real-user Core Web Vitals from the Chrome UX Report. It is fully free, needs no signup, and it is the only tool here that puts lab and field data side by side, for mobile and desktop.
The limits: the lab run emulates one mid-range phone on a heavily throttled connection, one run per click, from a Google data center. Treat the lab score as a lead and the field data as the verdict.
2. WebPageTest, the deepest free lab
WebPageTest, run by Catchpoint, is the most capable free lab tool. It loads your page in real browsers from real locations and produces a full request waterfall, a filmstrip of the page rendering, and a repeat view that shows what caching is worth. Connection profiles and run counts are configurable.
The depth costs learning curve. The free tier needs an account and caps monthly runs, and reading a waterfall is a skill of its own. For diagnosing one stubborn page, nothing free beats it.
3. Lighthouse in DevTools, for pages the web tools cannot reach
Lighthouse ships inside Chrome DevTools, which makes it the only entry on this list that can test localhost, a staging site behind a login, or a checkout flow mid-session. Open DevTools, pick the Lighthouse tab, choose mobile or desktop, run.
Results depend on your machine. A fast laptop flatters the score, a busy one tanks it, and single runs swing by several points. Use it during development; publish numbers from a controlled environment.
4. GTmetrix, friendly reports with history
GTmetrix runs Lighthouse under the hood and wraps it in the friendliest report of the group: a letter grade, clear summaries, and on a free account, history charts and scheduled re-tests. For watching a site drift over months without paying, it is the practical choice.
The free tier tests from one default location in Canada, which adds latency for European sites, and the letter grade invites score-chasing. Read the trend lines, not the letter.
5. Pingdom, the quick check for non-developers
Pingdom’s free speed test is the fastest path to a shareable result: enter a URL, pick a region, get load time, page size, request count, and a grade in seconds. It is the easiest report to hand to someone non-technical.
It predates Core Web Vitals and never adopted them, so its grading rewards old rules and misses what Google actually measures. Fine for a page-weight glance; not a basis for decisions.
Lab data vs field data, and why tools disagree
Lab data is one controlled load: one device, one connection, one moment, from one place. Field data is what real Chrome users experienced on your site over the last 28 days, aggregated at the 75th percentile. Google ranks with field data. Lab data exists to find causes, not to be the verdict.
This is also why two tools rarely print the same score: different hardware, different throttling, different locations, different run counts. A lab score is only comparable to itself over time, in the same tool with the same settings.
What Google considers good, at the 75th percentile of real visits:
| Core Web Vital | Good | Poor |
|---|---|---|
| Largest Contentful Paint (LCP) | ≤ 2.5 s | > 4.0 s |
| Interaction to Next Paint (INP) | ≤ 200 ms | > 500 ms |
| Cumulative Layout Shift (CLS) | ≤ 0.1 | > 0.25 |
If lab and field disagree, believe the field data about impact and use the lab run to locate the cause.
What slow pages cost
Speed is not a vanity metric. The evidence has been consistent for years:
- Visitors leave. Google’s mobile research found 53% of mobile visits are abandoned when a page takes longer than three seconds.
- Conversions follow milliseconds. The Google/Deloitte study “Milliseconds Make Millions” measured a roughly 8% lift in retail conversions from a 0.1 second speed improvement.
- Rankings notice. Core Web Vitals are a Google ranking signal. Not the biggest one, but the only one you control completely.
- Ads get more expensive. Google Ads factors landing page experience into Quality Score. A slow landing page pays more per click for the same position.
Most of the weight on a typical SMB page arrived one campaign at a time: a tag manager here, a chat widget there. The triage logic is the same as for tool sprawl: inventory what runs, keep what earns its cost, remove the rest.
What every free speed test still misses
Even the best free tools stop short of the questions that decide where to spend effort:
- What caching is worth. Almost no free report shows the repeat visit. WebPageTest can, if you know to ask. Without it, caching fixes look invisible in the score.
- Which third party costs what. Scores drop, but few reports name the tag manager, chat widget, or pixel responsible, in bytes and requests.
- Both devices in one run. Most tools default to one profile per test. Pages fail on mobile first.
- The score trap. A score is not a diagnosis. Teams that optimize the number end up with lazy-loaded everything and a page that still feels slow. Optimize the metrics that map to experience: LCP, INP, CLS.
FAQ
What is the best free website speed test? PageSpeed Insights for the standard read with field data, WebPageTest for deep diagnosis, Lighthouse in DevTools for staging and local work. No single free tool covers all three jobs.
Why do PageSpeed Insights and GTmetrix show different scores? Different lab conditions: hardware, throttling, location, and run count. Both are right about their own run. Compare a tool only with itself over time, and let field data settle arguments.
What is a good PageSpeed score? 90 and above shows green, but Google ranks with field data, not lab scores. A page at 75 with green Core Web Vitals in the field beats a page at 95 whose real visitors see red.
Does page speed affect SEO? Yes. Core Web Vitals are a ranking signal within page experience, and slow pages also lose the behavioral signals that matter more: visitors who bounce before the content paints do not convert, link, or return.
How often should I test website speed? After every release, and on a schedule once a month. Field data trails reality by up to 28 days, so a regression shows up in lab tests weeks before it shows up in the Chrome UX Report.
The gaps above are why we built our own free Website Speed Check: it loads your page eight times in a real browser (three cold runs each on emulated mobile and desktop, plus a repeat visit per profile), reports median Core Web Vitals, names the third parties by bytes and requests, shows what a warm cache is worth, pulls in Chrome UX Report field data when Google has it, and ranks the fixes worst first. Full disclosure, it is ours, and it needs no signup. Third-party tags are also a privacy problem; see The best free cookie scanners, ranked. And if the fixes the report names are work you want done rather than read about, that is what we do.