Oct 13, 2026
Best DNS Servers for Gaming (2026): Does DNS Actually Lower In-Game Ping?
Online gamers constantly search for the best DNS servers for gaming in hopes of slashing in-game ping, eliminating lag spikes...
Live DNS-over-HTTPS benchmark from your browser
Run a free DNS speed test online. Compare real DoH response times, check reliability, and find the fastest tested DNS server for your network.
Free · No installation · Real browser-based measurements
Ready to test your DNS performance
Compare DNS providers using your current network connection.
Your browser sends real DNS-over-HTTPS (RFC 8484) queries straight to each provider. Nothing is routed through our servers.
Each provider first gets unscored warm-up queries, then 10 scored rounds (15 in the Standard test). In every round all providers receive the same question, in a shuffled order. Every response is checked to make sure it really answers the question that was sent.
The time shown is the full request time seen by your browser, including network, encryption and resolver processing. Providers are ranked by median time. Only providers that answered at least 90% of queries are ranked. When two neighbouring results overlap statistically, they are marked as “similar”.
Results apply to this browser, network and moment. Speeds over plain DNS or DNS-over-TLS on your device can differ.
Most DNS benchmarks measure from a remote data centre. Fastest DNS sends real DNS-over-HTTPS queries from your browser, so the numbers reflect your connection to each resolver’s nearest point of presence.
GET /dns-query?dns=q80BAAABAAAAAAAABmdvb2dsZQNjb20AAAEAAQ host: cloudflare-dns.com accept: application/dns-message
Your browser builds a binary DNS query in RFC 1035 wireformat with a random ID, so no HTTP cache can answer it.
fetch() sends it to each resolver’s RFC 8484 endpoint as application/dns-message, exactly like a DoH-enabled browser.
The resolver answers from its edge cache or performs recursive resolution, then returns a binary DNS response.
performance.now() measures the full round trip. Warm-up queries are excluded so TLS handshakes don’t skew the score.
Each domain is served by different authoritative nameservers and sits in resolver caches differently. Testing one domain tells you about that domain; testing several tells you about the resolver.
Edit the list any time with “Hosts” in the tool — your choices are saved privately in this browser.
A cached answer comes straight from resolver memory. A cache miss walks the DNS hierarchy before answering.
A single slow Wi-Fi packet shouldn’t crown or sink a resolver. We rank by median latency, then check whether the gap to #1 is real.
2,000
resamples of each resolver’s
measurements per comparison
For each resolver we repeatedly resample its measurements, compute the difference in medians against the resolver ranked directly above it, and build a 95% confidence interval. If that interval includes zero, the two are labelled similar instead of declaring an arbitrary winner.
Privacy and family resolvers intentionally answer some domains differently. Those answers are still fast, valid DNS responses.
We time the response, not the answer — so blocking resolvers are scored fairly.
What DNS speed can and can’t change about your internet experience.
“A faster DNS server gives you higher download speeds”
DNS only translates domain names to IP addresses at the beginning. It does not carry data and has zero effect on bandwidth.
“1.1.1.1 is always the fastest resolver everywhere”
Routing depends on ISP peering and location. On many networks, Google or Quad9 answers faster.
“DNS-over-HTTPS hides all browsing history from ISPs”
DoH encrypts the query, but TLS SNI and destination IP addresses remain visible without ECH.
“Running a benchmark once gives an absolute winner”
DNS performance fluctuates with cache state, local Wi-Fi contention, and edge server load. Multiple rounds and median stats are required for reliable decisions.
Straight answers about DNS speed, privacy, and operating system configuration.
Fastest DNS creates RFC 1035 wireformat DNS binary queries directly inside your browser and dispatches them via fetch() over HTTPS (RFC 8484) to each provider’s secure DoH endpoint. We benchmark the complete round-trip time from your device to the resolver’s nearest point of presence.
Browsers enforce Cross-Origin Resource Sharing (CORS) security. Resolvers that do not send the Access-Control-Allow-Origin header cannot be read by web applications. This is an intentional security boundary of the web platform, not an indicator that the DNS server is slow or offline.
Both profiles start with unscored warm-up queries. The Quick test then runs 10 scored rounds of A (IPv4) lookups. The Standard test runs 15 scored rounds that alternate A (IPv4) and AAAA (IPv6) lookups, giving more samples for P95 tail latency and the statistical comparison. Administrators can adjust these numbers.
No. DNS only speeds up the initial lookup when opening a new page or connection. Once an IP address is known, download throughput and video streaming depend entirely on your ISP bandwidth and CDN routing.
No, DNS does not lower in-match tick-rate ping because games transmit live gameplay packets directly via UDP IP sockets once connected. However, a fast resolver can make lobby matchmaking, party connections and anti-cheat handshakes start sooner, because those steps look up hostnames first. Read our in-depth analysis in Best DNS Servers for Gaming (2026).
On Windows, run ipconfig /flushdns in Command Prompt. On macOS, run sudo dscacheutil -flushcache; sudo killall -HUP mDNSResponder in Terminal. On Linux, run sudo resolvectl flush-caches. For complete commands and browser cache clearing steps, see our guide on How to Flush DNS Cache on Windows, Mac, Linux & Chrome.
A DNS speed checker (also referred to as a DNS speed test or DNS benchmark) measures the time it takes for a Domain Name System (DNS) resolver to translate a human-readable hostname (such as fastestdns.online) into a machine-routable IP address (such as 104.21.49.201). Because internet applications must obtain a destination IP address before initiating transport handshakes, DNS lookup latency directly influences how quickly website connections and initial resource requests begin.
Our online DNS speed test measures browser-observed resolution time in milliseconds (ms) directly from your client device using DNS-over-HTTPS (DoH, RFC 8484). By dispatching concurrent queries to multiple public resolvers across identical domain hostnames, the tool measures observed latency, round-trip variance, and response reliability from your current network connection.
Running a DNS speed benchmark with our free online tool requires zero software installation, browser extensions, or administrative privileges. Follow these steps to measure resolver performance:
Unlike synthetic benchmarks that execute lookups from remote cloud datacenters (which measure data center peering rather than your local connection), our tool executes Real Speed Checks from Your Browser. Here is the technical methodology powering each measurement run:
fetch() requests with the application/dns-message content type straight to each resolver’s HTTPS endpoint. HTTP protocol negotiation (HTTP/2 or HTTP/3) is managed automatically by your browser’s native networking stack via ALPN.Each benchmark run generates detailed statistical telemetry. Use this reference guide to interpret your benchmark results accurately:
| Metric | Technical Meaning | Real-World Significance |
|---|---|---|
| Median (ms) | The 50th percentile response time across all valid completed queries. | The primary indicator of typical speed. Exactly half of the lookups were faster, and half were slower. |
| Average (ms) | The arithmetic mean of all completed query latencies. | Reflects overall consistency; a significant divergence between median and average points to sporadic latency spikes. |
| P95 Latency (ms) | The 95th percentile response time; 95% of queries completed faster than this value. | Measures tail latency and potential stalls during peak network load or cache misses. Note that in small sample sizes (10–15 queries), P95 serves as an indicative sample percentile rather than a high-precision distribution tail. |
| Minimum (ms) | The fastest single query recorded during the entire session. | Indicates best-case performance when both the local network path and edge resolver cache are fully warm. |
| Maximum (ms) | The slowest single completed query in the benchmark. | Highlights cold resolver lookups, upstream recursive query traversal, or momentary Wi-Fi retries. |
| Reliability (%) | The percentage of dispatched queries that returned a valid DNS answer. | Critical for stability. A resolver with low latency that drops queries will cause frustrating connection timeouts. Note that filtering resolvers may intentionally return NXDOMAIN or 0.0.0.0 for blocked domains based on policy. |
If you run two consecutive tests and observe minor variations in response times (e.g., 22 ms vs 28 ms), this is expected. Factors causing normal variance include:
There is no universal, scientifically absolute latency threshold because DNS response times depend heavily on physical connection type, geographic proximity to the nearest Anycast node, and transport protocol overhead. In general, browser-observed DoH response times fall into these illustrative contextual brackets:
Because network topology varies by Internet Service Provider and location, the optimal resolver for a user in Frankfurt or London may differ substantially from the best choice for someone in New York, Tokyo, or Sydney. Finding the right resolver requires a methodical approach:
A common misconception online is that changing your DNS server will increase your internet download or upload bandwidth. Here is how DNS actually interacts with internet performance:
DNS Does NOT Increase Bandwidth: Your maximum download capacity (such as downloading a large file or streaming video) is governed strictly by your physical network connection and ISP subscription speed tier. Changing your DNS resolver will not turn a 100 Mbps broadband plan into a 300 Mbps plan.
DNS Can Improve Perceived Browsing Snappiness: A modern website rarely loads from a single domain. Opening a typical media site or dashboard can trigger lookups across 30 to 80+ distinct hostnames for scripts, stylesheets, fonts, analytics, and third-party media. While modern web browsers resolve hostnames concurrently and reuse TCP/TLS connections via HTTP/2 and HTTP/3 multiplexing, reducing initial lookup delays across uncached domains helps browsers discover and initiate critical resource fetches sooner.
DNS and Online Gaming Ping: In online multiplayer gaming, in-game ping represents continuous UDP packet transit between your game client and the game server. Because DNS resolution occurs only once when establishing the initial lobby or matchmaking connection, changing DNS will not lower in-game ping. However, a reliable resolver prevents lobby connection dropouts and matchmaking lookup timeouts.
Public DNS resolvers offer free, high-performance alternatives to standard ISP-provided nameservers. Below is an overview of popular global providers (explore our full Public DNS Servers Directory for verified IPv4, IPv6, DoH, and DoT configurations):
Many online DNS tools display synthetic simulations or execute queries from remote cloud servers. Our DNS benchmark is designed for verifiable, client-centric accuracy:
A DNS speed checker is a network diagnostic tool that measures the response latency and reliability of Domain Name System resolvers. It dispatches DNS queries to multiple servers, times the round-trip interval in milliseconds, and identifies which resolver returns domain records most rapidly from your network location.
When you click Check DNS Speed, your browser sends real DNS-over-HTTPS queries for test domains directly to each DNS provider and measures how long each complete, valid response takes. Each provider first receives unscored warm-up queries, followed by 10 scored rounds (15 in the Standard profile). Providers are ranked by median response time, and only resolvers answering at least 90% of queries are ranked.
Web browsers enforce Cross-Origin Resource Sharing (CORS) security rules. A browser can only read a DNS-over-HTTPS response if the resolver includes an Access-Control-Allow-Origin: * HTTP header. Resolvers such as Quad9 and NextDNS do not currently include this header on their public DoH endpoints. Rather than presenting simulated numbers, our tool transparently marks them as “not tested from this browser”. You can review their configurations in our Public DNS Directory and compare server-side probe data.
Latency is important, but it is not the only consideration. You should also evaluate a provider’s privacy commitments (whether they retain IP query logs or monetize telemetry), threat filtering features (such as automated malware blocking), and reliability history. A resolver with 5 ms higher latency that actively blocks malicious phishing domains is often a superior everyday choice.
For most global users, Cloudflare (1.1.1.1) and Google Public DNS (8.8.8.8) frequently deliver the lowest latency due to their extensive Anycast edge infrastructure. However, in specific metropolitan regions, regional providers like Control D, Quad9, or local ISP resolvers may deliver comparable or lower response times. Running our live browser benchmark determines the fastest option for your specific ISP path.
Yes. Our DNS speed checker is fully responsive and runs in any modern mobile browser (Chrome, Safari, Firefox, Edge) on Android, iPhone, and iPad. Simply access the site on your mobile device over Wi-Fi or cellular data and tap Check DNS Speed.
Traditional ISP resolvers operate over unencrypted UDP port 53. Because web browsers run inside a sandboxed environment and cannot dispatch raw UDP sockets or inspect OS network adapter settings, your browser cannot directly benchmark your default ISP DNS unless your ISP exposes a public DNS-over-HTTPS endpoint with open CORS headers. To determine which resolver your operating system is currently using, consult our step-by-step What Is My DNS Server? guide.
Yes. Click the Custom DoH Servers button above the leaderboard. Enter a friendly name and a valid HTTPS DoH endpoint (such as your private Pi-hole, AdGuard Home, or NextDNS URL). As long as your endpoint is served over valid HTTPS and returns standard CORS headers, our engine will benchmark it alongside public providers. Note that browser Private Network Access rules may restrict public websites from directly querying local RFC 1918 private IP addresses.
A DNS leak occurs when your computer or mobile device accidentally routes DNS requests through your local ISP or unencrypted default gateway instead of the encrypted VPN tunnel, exposing the hostnames of websites you visit. In 2025, modern encrypted operating system features and split-tunneling configurations make auditing for leaks essential. To test for leaks: connect to your VPN, verify your external IP has changed, and review our dedicated DNS Leak Test Guide to confirm your queries are handled exclusively by your VPN provider's private recursive resolvers.
DNS propagation is not an active “broadcast” across the globe; rather, it is the passive expiration of cached records across thousands of independent recursive resolvers. The primary cause of propagation delay is the Time-to-Live (TTL) setting configured on the previous DNS records. If your old nameserver record had an 86,400-second (24-hour) TTL, recursive resolvers worldwide are permitted to serve that stale cached record for up to 24 hours before querying your new authoritative nameservers. To minimize downtime before updating nameservers, lower your TTL to 300 seconds (5 minutes) at least 24 hours in advance.
Comparing top public resolvers like Cloudflare (1.1.1.1) and Google (8.8.8.8) helps identify the lowest latency path for your specific ISP routing. Leading free benchmarking tools include Fastest DNS (our browser-based DoH tester measuring real client latency without software installations), GRC DNS Benchmark (a legacy Windows utility for local UDP testing), and DNSPerf (which aggregates synthetic performance metrics from worldwide data centers). For real-world everyday browsing, running tests directly from your actual client device provides the most accurate and actionable measurement.
For newly registered domain names that have never had existing cached records on recursive resolvers, DNS propagation is usually very rapid—often resolving globally within 15 minutes to 2 hours. However, registry-level updates at the Top-Level Domain (TLD) nameservers (.com, .org, .io) and variations in regional ISP resolver caching policies can occasionally extend full worldwide propagation up to 24–48 hours.
If your network diagnostics show Google DNS IPs (8.8.8.8 / 8.8.4.4) instead of your intended ISP or VPN resolver, your system is likely experiencing one of three common configuration overrides: First, modern web browsers (Chrome, Edge, Firefox) often enable “Secure DNS” by default, which overrides operating system network settings to route queries directly to Google or Cloudflare over DoH. Disable or adjust Secure DNS in your browser settings. Second, your home router may have Google DNS configured as its WAN fallback. Third, your VPN client may lack IPv6 leak protection, causing IPv6 DNS traffic to leak outside the tunnel. Consult our DNS Leak Test Guide for platform-specific mitigation steps.
Explore our companion DNS tools and educational tutorials to further optimize your network performance and digital privacy:
Practical, fact-checked writing on DNS performance and privacy from our engineering team.
Oct 13, 2026
Online gamers constantly search for the best DNS servers for gaming in hopes of slashing in-game ping, eliminating lag spikes...
Oct 13, 2026
When configuring a custom public DNS resolver to replace slow or unreliable internet service provider (ISP) defaults, two tit...
Oct 13, 2026
Whenever you switch to a faster DNS resolver, migrate a website to a new hosting server, or encounter dreaded browser errors...