Looking glass
Choose the location closest to your users
Pick where your users are and see how far each of our 10 data centers is from them. Then deploy on monthly billing and measure the real route yourself.
Closest location: London (~5 ms)
Figures marked measured are medians of real measurements (WonderNetwork, September 2026). The others are estimated: fibre at about two-thirds of the speed of light, 1.25 × the straight-line distance, plus 5 ms for equipment and local routing — typically within ±5 ms of reality.
Measure your routeLatency from your users to each location
Measure it yourself
Check the real route from your network
Billing is monthly with no contract: deploy in the location you picked, run these two tests in the first hours, and simply do not renew if the route does not suit you.
What to run on day one
# Latency and loss, hop by hop
$ mtr -rwzc 100 <server-ip>
# Throughput: start "iperf3 -s" on your server, then from your side
$ iperf3 -c <server-ip> -P 8 -t 20Reading the results
- Latency close to the estimate above means a direct route; much higher usually means traffic detours through another city.
- Loss should be 0% on the last hop. Loss on intermediate hops only is usually ICMP rate limiting, not a problem.
- Throughput from a single connection is limited by your own line and by distance; parallel downloads show what the port can do.