Measuring…
The verdict for this game appears once the first full second of packets is in.
Comparing with other tests to this server…
Measuring…
The verdict for this game appears once the first full second of packets is in.
Comparing with other tests to this server…
Are you sure you want to delete all results?
| Date & Time | Duration | Avg Ping | Avg Jitter | Packet Loss | Actions |
|---|
Pals frozen mid-task at your base? Kicked from a friend's world for no visible reason?
Palworld has two multiplayer modes that behave so differently that advice for one is often wrong for the other. Invite-code co-op runs the world on one player's machine for a small group; a dedicated server runs it on a machine that exists to do nothing else and supports far more players. Which one you are on decides whether your problem is somebody's home connection, your own, or a server struggling with a base full of working pals.
What this test sends: 30 packets per second of 350 bytes, matching Palworld rather than a generic ping. estimated
Pocketpair publishes no tick rate; dedicated servers expose no rate setting. Unreal's default replication is 30 Hz.
Run the test above for at least 60 seconds. Here is what each number means for this game specifically.
Palworld responds to a broken connection by removing you from the session, usually without much explanation. Sustained loss on this chart is the most common reason, and unlike a server problem it will show up here consistently rather than only at the moment you are dropped.
Capture attempts, pal commands and your own shots are all confirmed by the host. Uneven delivery produces the pattern where a sphere connects visually but the capture resolves a beat later, or where a pal ignores an order and then obeys it twice. That is delivery timing, not the game deciding to ignore you.
There is no region to select. In co-op you are connecting to the host player, and on a dedicated server you are connecting to wherever that box physically is. A community server hosted on another continent will read high here and there is nothing local to do about it except choose a nearer one.
What you see in game, and which number on this page explains it.
| What you see in Palworld | Look at | What it means |
|---|---|---|
| Dropped from the session repeatedly | loss | Sustained loss broke the connection to the host. |
| Capture resolves a beat after the sphere lands | jitter | The host's confirmation arrived late, not the capture failing. |
| Pals frozen, players moving normally | server | The host machine is busy simulating the base, not the network. |
| High ping with a clean chart | ping | The world lives on a distant machine. There is no region to switch. |
Palworld's two ways of playing together are not two front ends onto the same system. Invite-code co-op takes a world that lives on one player's machine and lets a small group into it, with that machine doing all the simulation while also running the game for its owner. A dedicated server takes the same world and puts it on a machine whose only job is to be the authority, which is why it can hold far more players.
That distinction decides where your problems come from. In co-op, the host is a home connection with a home router and a household sharing the line, and everything that happens to it happens to the group. On a dedicated server, the host connection is out of the picture and what remains is your own path to a specific machine - which is exactly what this page measures.
The base building adds a second, less obvious load. Pals assigned to work continue doing so as the world runs, and the machine hosting the world is the one calculating all of it. A large base with many pals is real, ongoing processing that has nothing to do with the network, and it produces a distinctive symptom: workers freezing or acting stupidly while players still move and shoot normally.
Keeping those two apart is worth the effort because they need opposite responses. Pals stalling while the chart here is clean means the host needs a better machine or a smaller base, and no amount of network tuning will touch it. Players warping and disconnecting while the chart shows loss means the opposite, and the fix is on the connection rather than in the game.
If the chart above shows packet loss or high jitter, work through these in order.
The Symptom: Advice that works for others does nothing for you.
Invite-code co-op and a dedicated server have different player limits, different configuration and different failure modes. In co-op the host's home connection is the infrastructure; on a dedicated server it is not, and the settings that matter live in the server's configuration file rather than anywhere in the game's menus.
The Symptom: The world is only smooth when one person hosts.
In co-op, one player's upload carries the world for everyone else. Wireless bursts on that machine become everyone's stutter and everyone's disconnects. If you are the one with the world save, a cable is the highest-value change in the whole group.
The Symptom: Pals stop working, but movement is fine.
Base pals are simulated by the host, and a large base with many pals is real work for that machine. When pals freeze while players move normally, that is the host's processor rather than the network - and this chart being clean is how you confirm it.
The Symptom: Disconnects when something else is downloading.
The game's continuous traffic is small and easily starved by a saturated link, which produces exactly the pattern of being dropped whenever the household gets busy. Router QoS that prioritises interactive packets removes it.
Sustained packet loss is the usual cause you can act on. Run this test for a few minutes: consistent red here means your line, while a clean chart during repeated drops points at the host or the server instead.
For anything beyond a small group, yes. It supports far more players, it does not require the world's owner to be online, and it stops one household's connection from being the whole group's bottleneck. The trade is that you have to run and configure it yourself.
Base pals are simulated by whichever machine hosts the world. A large base is genuine processing load, and pals stalling while players still move normally is a sign of the host being busy rather than of a network problem.
Not in the sense that a game with official servers offers. You connect to a specific machine - a friend's, or a community server at whatever address it has - so the latency you get is a property of that choice.