DNS Resolver Speed Race
Time the public DNS resolvers on names none of them have cached.
Run the test
Each sample is a real HTTP request to the nearest Cloudflare edge, with the response read in full. It is not an ICMP ping: it includes the server’s own handling time and stops at the edge rather than reaching any particular destination.
Where this comes from
Each resolver is asked to resolve five names that have never existed, one after another, and the median is taken. The choice of names is the whole methodology: the first lookup of a name is a real resolution, walking the hierarchy from the root, while every lookup after it is served from cache and reflects nothing but network distance. A comparison that asks each resolver about a popular domain is timing how fast they read their own memory, which is why those comparisons produce suspiciously similar numbers.
What the results mean
- Median response
- The middle of five uncached resolutions. Half the queries were faster and half slower.
- Spread
- Between the fastest and slowest resolver. Under about 20 ms is not noticeable in browsing; more is, because a busy page resolves a dozen domains.
- No answer
- A resolver that failed repeatedly. Some networks block DNS-over-HTTPS endpoints outright, which is itself worth knowing.
- Uncached names
- Names invented for this test, so no resolver can have them cached. It is what makes the comparison mean anything.
Common problems and fixes
- All three are within a few milliseconds
- That is the normal result on a well-connected line — they are all anycast networks with edges near you. Choose on filtering and privacy policy instead.
- One resolver never answers
- Some networks and some routers block specific DNS-over-HTTPS endpoints. That is a finding about the network you are on rather than about the resolver.
- Changing my resolver did not speed up browsing
- Expected. DNS is a few milliseconds of a page load that takes hundreds, and most names on a page you visit often are already cached locally.
Frequently asked questions
Does changing DNS resolver make browsing faster?
Marginally, and much less than the claims suggest. Resolution happens once per domain and is then cached locally, so it contributes a few milliseconds to a page load that takes hundreds. Where a change is genuinely felt is when the current resolver is badly overloaded, slow to reach, or filtering things you did not ask it to.
Why do resolver comparisons all show similar numbers?
Because most of them query popular domains that every resolver already has cached, which measures a memory lookup and a network hop rather than a resolution. Asking for a name that has never been resolved is a completely different measurement, and it is the one this uses.
Which DNS resolver should I use?
For most people the difference in speed is not the deciding factor — the spread here is usually small. Filtering behaviour, logging policy and whether the provider blocks anything by default matter considerably more, and those are choices rather than measurements.
Put this on your own site
Free to embed, no attribution required beyond the source link the frame carries itself. It runs entirely in your visitor's browser, sets no cookies and loads no third-party script.
<iframe src="https://runthetests.com/embed/dns-resolver-speed-test/" width="100%" height="560" style="border:1px solid #e5e5e5;border-radius:8px" title="DNS Resolver Speed Race" loading="lazy"></iframe>
Preview it at https://runthetests.com/embed/dns-resolver-speed-test/. Embedded pages are marked noindex, so yours stays the canonical copy — not this one.