But 73 of those “servers” use Smart Routing, meaning you are actually connecting to another VPN server but your IP appears to be from the fake VPN server you are connecting to .
What are more people looking for when they’re comparing the location count of VPNs? Do people care about where their IP address appears to be, or do they care about their physical proximity to the server?
For some it may be both, but my guess is that most people care about the first thing more. If that’s the case then Proton’s solution here still gives an accurate country count, because that is indeed the number of countries you can appear to be located within based on your IP.
I just counted and 73 servers out of their 113 are in fact not servers, but rather use Smart Routing to redirect to other countries, mostly France. This is highly misleading. Someone in let’s say Africa will look at their website, and say “Oh great, they have nearby servers”, while in fact, it is on another continent. ( I take this example cause Africa has very few VPN servers).
Plus, even PG says choosing a nearby server is best for latency.
I also want to point out that the countries count is supposed to be an indicator of the size of the VPN infrastructure.
Not sure why they don’t show this information on the website, but they still do have the most amount of servers in the most amount of countries when compared to IVPN and Mullvad.
This is false. Mullvad has real/physical servers in 47 countries, compared to Proton’s 40. In raw number of servers, they do have more. But not in terms of countries.
Strong disagree. Most people DO care about latency and speed (ask me how I know), and fake locations don’t check this box.
For those who “want an IP in a country” virtual locations don’t solve this either, since its trivially obvious that its just geo-ip spoofing, that is revealed with a basic traceroute. In vast majority of cases “virtual location IPs” are not good enough to access most geo-restricted websites.
So if they don’t provide any performance benefits, and don’t unblock most geo-restricted websites, what is the point of them?
The ONLY point is to have a bigger number that VPN review sites can use to say “this is why provider X is better”. That’s it.
I think only Secure Core servers are Proton’s own servers. Rest are virtuals or rented ones.
Proton VPN operates 113 Secure Core servers in 65 countries, which are available to Plus plan users.
Can’t say anything about Windscribe’s servers because I know many of the IPs don’t belong to the countries that it is showing. Yes, whatismyip is showing IP belongs to that country, like Turkey, but services which are checking the IPs are saying otherwise, e.g. Google, PSN, Disney, etc.
This doesn’t seem to be true actually. When I was testing Proton VPN a while ago, I noticed that many of the servers had the sections “Organization” and “ISP Name” marked as Proton or Proton VPN in ip.me. This is a site that Proton owns. So, I asked about this from their support and they confirmed that these were in fact their owned servers. I was only using servers from the Netherlands and Finland during my testing period, and both of these locations had also Proton-owned servers, so it is very likely that Proton also has these servers elsewhere.
I think this topic deserves its own thread. For this particular case, though, we have enough information. According to Proton, they serve 122 countries, of which 67 countries is served by Smart Routing. So they have real servers in 55 countries as of today.
We can fix this with a simple warning card that states something like this:
Proton VPN uses Smart Routing to offer VPN services in 67 countries. You will experience worse performance with these servers. Therefore, we recommend against using them if possible.
We should also update the number of countries that iVPN serves, which is 41 at the moment.