Server locations and route distribution

A point-to-point dedicated channel that never crosses the public internet backbone.

How it works

Traffic enters the dedicated channel at a local access point, runs on an isolated link the whole way, and only returns to the public internet at the two endpoints. Fewer routing hops, a fixed path, and no rerouting when the public network congests.

Best for

Peak-hour streaming, long AI tool sessions, remote work and any task that has to stay online. Pick it first when dropped connections are the concern.

Cost

The scarcest and most expensive of the three, so it is deployed only in popular regions and major exit cities, and there are fewer of these routes than of the other two types.

Connect to the nearest entry point first, then follow an optimised path to the target exit.

How it works

Your traffic first connects to a nearby entry node, then travels along an optimised path to the target exit. Entry and exit are separated, which sidesteps the most congested stretch of the public internet.

Best for

Everyday browsing, cross-region access and long-term use where value matters most. It covers the most countries, so even less common regions usually have an exit here.

Cost

Sitting between dedicated and direct, this is the best balance of coverage and price and the default starting point for most scenarios.

Straight from your location to the local exit — the shortest path.

How it works

Your local network connects directly to the exit node with no extra relay hop. Shortest path, fewest hops, and the most dependent on the quality of your own connection.

Best for

Reaching services local to the exit city, lightweight lookups, and acting as a backup for dedicated and relay routes. Switch to direct first when a main route is under maintenance.

Cost

The cheapest of the three and the most numerous, so it is worth keeping one or two as emergency exits.