Guide
How to test your VPN speed properly
A VPN speed result on its own means nothing. Most "my VPN is slow" reports turn out to be measured the wrong way: a test server the site chose automatically, a busy Wi-Fi channel, or a connection that was never fast to that country in the first place. Five minutes of method gives you a number you can act on.
Four steps to a number you can trust
Do these in order. The whole method rests on one idea: measure the route your traffic will actually take, not the one your speed test would prefer.
-
Start with the location, not the app
Decide which country you will connect through, because everything else follows from it. A VPN sends your traffic to that country before it goes anywhere else, so that route, and not your local connection, is what you are about to measure.
-
Test that route before you connect
Disconnect the VPN and run a speed test, but choose the test server by hand, in the country you picked. On Speedtest.net that is the Change Server link next to the server name: search for a city in that country and select it. Connecting through the United States means testing against a United States server. Never leave it on automatic, because the site picks the server nearest you, sometimes one inside your own ISP, which flatters the result and measures nothing like the path a VPN takes. The number you get is your ceiling. If your connection to that country is already slow, no VPN can make it faster, and you have your answer without going further.
-
Connect and repeat, changing nothing else
Connect to that country and run the test again against the same test server. Same device, same cable or the same spot on the Wi-Fi, same time of day, with nothing streaming, syncing or updating anywhere in the house. Wi-Fi on a crowded channel swings more than any VPN does.
-
Compare like for like
Now the two numbers describe the same route, so the difference between them is the VPN and nothing else. A single run proves nothing: repeat three times and take the middle result. Ping will always rise, because your traffic reaches the server before it reaches the test.
Where to test
Only the first of these lets you choose where you are testing to, which is what this method needs. The other two are worth a second opinion once you have your numbers. They are run by other people; we do not control them and do not track what you do there.
-
Speedtest.net (opens in new tab)
The one to use here, because Change Server lets you pick a city in your VPN's country and keep testing against it. Use the same test server for both measurements, or the comparison means nothing.
-
Cloudflare speed test (opens in new tab)
Picks its own location, so it cannot test a specific country. Useful for what it measures instead: loaded latency and jitter, which decide whether calls and games feel smooth.
-
Fast.com (opens in new tab)
Netflix's tester, also automatic. Useful for one specific question: whether streaming from that network is the thing being slowed down.
What a normal result looks like
Every VPN costs something. Your traffic is encrypted, wrapped, and sent through one more machine before it reaches the site you asked for, and none of that is free. On a nearby server, a healthy connection keeps most of its download speed and gains some ping. Losing a slice of throughput is normal. Losing almost all of it is a signal, and usually points at one of the causes below rather than at the VPN itself.
-
Distance shows up as ping
Ping rises with physical distance and never falls below it. If ping tripled, you are measuring geography, not the VPN. Reconnect to a nearer country and test again.
-
The route is the ceiling
A VPN cannot beat the connection underneath it. If step two was already slow to that country, that is your ceiling, and the fix is a nearer location rather than a different VPN.
-
One slow site is not a slow VPN
If a speed test looks fine but one service crawls, that service is throttling or routing you differently. Test a second site before blaming the tunnel.
What you can change in the app
Four settings move the needle, and each one is a trade rather than a free win.
-
Try another protocol
UDP is the lightweight option and usually the fastest, though some networks restrict it. TCP gets through where UDP does not, at a cost. QUIC is faster again where it is allowed. The protocol is a per-connection choice, so switching costs nothing but a reconnect.
-
Turn Cloak Mode off when nothing is blocking you
Cloaked traffic carries more overhead than a direct tunnel, so it is a little slower and uses slightly more battery. That is the trade: raw speed when nothing is in your way, a connection that still works when something is. On an open network you do not need it.
-
Turn Smart Proxies off when the server is reachable
A proxy adds a hop between you and the server, and a public proxy is often the slowest link in the chain. It exists to get you in where the server is blocked outright, not to make a working connection faster.
-
Keep heavy traffic outside the tunnel
Split tunneling lets a backup, a game or a local device connect directly while everything else stays protected. Traffic that never enters the tunnel cannot be slowed by it, and it frees capacity for what does.
Free servers and Premium servers
VpnHood! does not throttle anyone. The server software can cap speed per user, and on our free servers it is not set. What differs is capacity: free servers carry many more people, so your share of one rises and falls with how busy it is. Premium servers are more powerful and shared only among premium users, which is why they stay fast at peak hours. If you run your own server with VpnHood! CLIENT, the same steps apply and the machine's uplink is your ceiling, so the fix is a bigger server rather than a different plan.
FAQs
Frequently Asked Questions
Let's find the answers to your questions.
