Our own
network.
RedVPN does not rent its way onto the internet. We run our own network, with 100 Gbps ports at two of the largest Internet Exchanges in the world, in Los Angeles and Frankfurt, and direct peering with the networks your traffic actually goes to.
Where we peer
Each of our two regions sits at a major Internet Exchange: the places where networks hand traffic to one another directly. One port at an exchange reaches hundreds of networks in a single step, with no transit provider in between.
Los Angeles
- Exchange
- Any2 West
- Facility
- CoreSite LA2
- Port
- 100 Gbps, IPv4 and IPv6
Any2 is one of the largest exchanges in the United States and the main one on the west coast. Our own hardware in Los Angeles connects here, so traffic to the Americas, the Pacific and the big US content networks takes the short path.
Frankfurt
- Exchange
- DE-CIX Frankfurt
- Facility
- Digital Realty FRA1-27
- Port
- 100 Gbps, IPv4 and IPv6
DE-CIX Frankfurt is one of the largest Internet Exchanges in the world. Our Frankfurt point of presence connects here and serves the European side of the network, so traffic to Europe, the Middle East and Asia takes the short path too.
One hop to the big networks
Beyond the exchanges, we peer directly with Meta, Google and other large content networks, at 100 Gbps. Most of what a phone does online, video, calls, maps, mail, app updates, ends up on a handful of networks, and those are the ones we connect to directly.
With peering, a packet leaves our router, crosses the exchange and arrives at the destination network: one hop. Without peering it would pass through two or three transit providers first, and each one is a place to add delay, throttle or watch.
For you that means one hop: from our server straight into Meta's or Google's network, instead of a path across two or three transit providers we do not control. Lower latency, so calls connect and video starts without the wait. And no middle networks on the way to throttle the traffic or watch it.
How it is built.
Three layers between your phone and the open internet: an entry point on a cloud address, a private tunnel, and a server we own. Here is the shape of it, without the addresses.
What the network you are on sees is a connection to a cloud address. What carries your traffic is a server we own.
- 1Entry points
Cloud fronts in the United States and in Frankfurt: the address your device connects to. They never hold your traffic in the clear.
- 2Private tunnels
Each entry point forwards over a private, encrypted tunnel to the back-ends. The back-ends have no public face of their own; only the entry points reach them.
- 3Our own servers
Traffic leaves from hardware we run in Los Angeles and from dedicated servers in Falkenstein, Germany. No rented exit nodes, no third-party VPN backbone.
Two transports
The first is standard WireGuard, carried over TCP so it passes networks that drop unfamiliar UDP. The second is a transport we built ourselves for networks where the first one struggles. The app tries what works on the network you are on and switches without asking you.
Configuration is signed and fetched from our own domains, and the app connects only to our own servers. There is no marketplace of third-party nodes behind it.
What this means for you
On an airport or hotel network, the operator sees one encrypted connection to a cloud address and nothing else: no sites, no apps, no messages.
We keep no account for the trial, because there is none: the 3-day start needs no sign-up. Premium is tied to the email or store account that bought it, and nothing more.
Two transports
WireGuard over TCP, and one of our own for unreliable networks.
Signed configuration
Fetched from our own domains; the app refuses anything unsigned.
Only our servers
No third-party exits. We name the cities and we mean them.
Linux with checksums
A static binary and SHA-256 sums, so you can verify the client you run.
What we do not publish
No addresses, hostnames or server counts, and nothing about the internals of the second transport. The exchanges and facilities above are on our public PeeringDB record already; the entry points and back-ends are not. Every detail we print is read by more than our users, and the people this network exists for pay for that, not us.
Peer with us
For network engineers: we are AS401418, Global Peering Alliance, with an open peering policy, ports at Any2 West and DE-CIX Frankfurt, and IPv4 and IPv6 on both.
Our PeeringDB record: peeringdb.com/asn/401418
Peering requests: noc@redteam.net