The internet wasn’t built for
competitive play.
So we built a smaller one.

ZeroLag started with a familiar bug report: a tier-1 CS player on 1 Gbit symmetrical fiber, losing rounds to peeker’s advantage. The line tested clean. The router was fine. The ISP swore everything was nominal.

We pulled traceroutes for a week. The packets were taking the long way around — Helsinki to Frankfurt via London via Amsterdam because that’s how the transit contracts settled at 2 a.m. when the BGP tables last reconverged. The internet was doing exactly what it was designed to do, and it was wrecking his ping.

The public internet routes for cost and reachability. It does not route for latency. If you play competitively, that distinction is the whole game.

We started small. Two POPs, a wavelength between them, and a thin client that re-routed game traffic onto our path. The first round of beta testers gained 14 ms on average. A few gained 40.

From there it was a build problem. We added POPs near the cities where matchmaking servers actually live — Frankfurt, Stockholm, Warsaw, Madrid, Dallas, Ashburn, Singapore. We wrote a route optimizer that picks the leg combination with the lowest measured jitter for your destination, refreshed every few seconds.

Today the network runs 48 POPs on dedicated fiber, peered directly with the game publishers we care about. No transit hops we don’t own. No surprises at 2 a.m.

We built ZeroLag because we play the games we’re routing for. That’s also why pricing works the way it does: a 3-day free trial, no card required, and if your measured ping doesn’t improve on your games, you walk. The product has to earn the milliseconds.

Everything else — the status page, the route guide, the FAQ — exists so you can verify that for yourself. We’re a small team. We answer our own support tickets. We count frames too.