Palworld hosting
Palworld servers that don't
need 24 GiB to stay up.
Palworld's memory use creeps upward the longer it runs. Most hosts answer by selling you 24 GiB. We restart your server on a schedule instead — every four hours by default, with warnings in-game first — so it runs well on far less.
What you get
Scheduled restarts
Every four hours by default, with players warned beforehand. Change the schedule or turn it off; we'll tell you how much more memory that needs.
Every setting, explained
Every PalWorldSettings.ini option as a field with a note on what it does. The admin password is generated for you, and you can change it.
Live player count
The dashboard shows who's online, read from Palworld's own server API.
Full file access
SFTP access to every file on your server: edit configs, upload mods, download your saves whenever you like.
A console that remembers
A live console in the browser. Leave and come back and everything since the server started is still there.
Reserved memory
The memory you pay for is held for your server alone and never sold to anyone else.
How much do I need?
Palworld grows with your bases more than your player count: a month-old world with five players can use more than a new one with ten. These sizes assume scheduled restarts are on.
| World | Memory | CPU | Disk included | Monthly |
|---|---|---|---|---|
| Small group, new world | 4 GiB | 2 shared cores | 30 GB | $14.50 |
| Recommended for most groups | 8 GiB | 3 shared cores | 60 GB | $25.75 |
| Big group, months-old bases | 12 GiB | 4 shared cores | 90 GB | $37.00 |
With scheduled restarts off, plan on roughly half as much memory again. See pricing for every option.
Good to know
- Do I need to own Palworld?
- Every player needs their own copy of the game. The dedicated server software is downloaded from Steam when your server is installed.
- Can I bring my existing world?
- Yes. Upload your save over SFTP before first start.
- Where are the servers?
- In Texas, United States. Players across North America typically connect with low latency.