Pay for twenty-four months and save 17% or moreOn every plan, pick the 24-month term when you order
Skip to main content

Enterprise Plan

Enterprise is built for a network behind one address, not for a single server that happens to draw a crowd. Sixteen cores mean a busy lobby does not slow the rest down, and the memory is wide enough to give each server its own budget instead of having them fight over the same few gigabytes.

A block of emerald, the icon of the Enterprise Minecraft server plan
  • 16 vCPU
  • 64 GB RAM
  • 600 GB SSD
What the Enterprise plan costs on each billing term.
TermPrice per monthSetup fee
24 months€67.99None
12 months€76.49None
6 months€80.79None
Monthly€84.99€2.99 once

Order the Enterprise plan

When one machine holds the whole project

Enterprise is 64 GB with 16 vCPU and 300 GB NVMe, and it exists for the network that has outgrown its own success: five or six sub-servers, a proxy, a database, and a staff team that runs it as a schedule rather than a hobby.

At this point the plan is not a Minecraft question any more but a capacity question. You are dividing memory between servers on purpose, and the sixteen cores decide how many of them can be busy at the same second.

What people actually run here

A modded network where each pack gets its own instance. A public survival server with a hundred regulars and a map render running beside it. A creative build server with hundreds of loaded plots.

Storage is 300 GB NVMe, which is what a network of world folders plus a backup history really costs once you stop deleting things to make room.

The price, and the one step above

Enterprise costs EUR 84.99 per month, EUR 67.99 on twenty-four months. Above it there is exactly one fixed plan left, Dedicated at 96 GB, and after that you build your own size on the slider.

What a network needs besides memory

At 64 GB the thing that breaks is rarely the machine. It is a database that nobody watches, a staff account with too many permissions, or a backup that has been silently failing since June.

Networks live on the boring parts: one place to see the logs, a restart schedule people know about, and a rule about who may install a plugin. All three are cheaper than the plan.

Sub-servers, and what they each need

A lobby needs almost nothing and gets almost everything wrong when it is starved. A modded instance wants a fixed slice that does not move. A creative build server wants disk more than memory, because loaded plots and schematics stack up faster than anything else on a network.

Write that division down somewhere. On 16 vCPU you can afford several of them being busy at once, but only if nobody quietly gives one instance twice what it was meant to have.

Moving a live network here

A migration at this size is planned in the order players notice: the database first, then the worlds, then the address. Keep the old machine running until the new one has answered a full evening, and switch the DNS when nobody is on rather than when it is convenient for you.

We keep the old server reachable during that window, so a rollback is a change of address rather than a restore.

Numbers worth watching every week

Tick time, memory after a restart, disk free, and the number of chunks loaded at peak. Those four say more about a network than any dashboard full of graphs, and they are the ones that predict the evening that goes wrong.

The Copilot reads the same logs and can tell you which of the four moved this week, which is faster than reading them yourself on 16 vCPU worth of instances.

Splitting 64 GB across a real network

A shape that works: lobby on two, survival on sixteen, skyblock on twelve, a modded instance on sixteen, a creative world on eight, and the remainder untouched for backups, the proxy and the database. That is 64 GB carrying five servers without any of them living on the edge.

The temptation is to hand out the last gigabytes to the biggest instance. Resist it; the free memory is what keeps a backup from taking the network down with it.

Questions at network scale

Can I run a database beside the servers? Yes, and on this size it is normal rather than exotic. Can staff get their own logins without the file manager? Yes, the panel separates that. Can we move one instance to another region later? Yes, and that is a planned job rather than a button.

Is there a plan above this one? Dedicated at 96 GB, and after that you build the size yourself.

When staff outnumber the sysadmin

Networks that reach 64 GB usually have more moderators than technical people, and that is where things quietly break: a plugin installed by somebody helpful, a config edited during an argument, a restart nobody announced.

The fix is boring and works: one person owns the file manager, everybody else gets the tools they actually need, and every change to a live instance happens outside peak hours.

The same step on the other editions

Kingdom at 32 GB is the Bedrock plan here. Very few Bedrock communities need this much, and the ones that do are usually running many small worlds rather than one enormous one.

On crossplay this rung exists too, and it is bought by networks who would rather translate once, centrally, than explain to half their players why they cannot join.

How to tell whether Enterprise is enough

Read the free memory across all instances at peak, and read it as a single number. A network that never drops below a few free gigabytes is healthy; one that touches zero on Saturdays is one plugin away from an evening nobody enjoys.

That single number is also the one worth putting in front of your staff team, because it makes upgrades a fact rather than an argument.

An instance list that works

Proxy, lobby, survival, skyblock, creative, events, and a staging instance nobody plays on. Seven names, one machine of 64 GB, and the seventh is the one that saves your Saturdays: every plugin update and every config change lands there first.

Networks that skip the staging instance test on their players instead, and they find out on the busiest evening of the week.

Costs that are not the plan

A network of this size usually pays for a domain, a Discord boost, a couple of premium plugins and someone's time. Against EUR 67.99 a month for the machine, those are the numbers that decide whether the project is sustainable, and they are the ones people forget when they compare hosts.

We can only be honest about our part of it: no setup fee on the longer terms, the price you see includes VAT, and nothing on the feature list costs extra later.

Moving off Enterprise, in either direction

Networks shrink as well as grow, usually after a pack ends or a season closes. Dropping to Premium or Ultimate is the same job as growing into this plan and just as uneventful: the worlds move, the address stays.

Growing instead means Dedicated at 96 GB, and that is worth a conversation first, because at that point splitting across two machines is sometimes the better answer than one larger one.

How it compares

Enterprise has 16 GB more memory than Ultimate, for €28.00 more a month on the twenty-four month term. If this gets tight, Dedicated is the next step up: 32 GB more for €20.00 more a month.

What runs comfortably on it

Enterprise has 64 GB, and this is what the RAM guide says fits in that: RAM guide.

Workloads the Enterprise plan's memory covers.
What you runMemory the guide lists
Vanilla 1-5 players2-4 GB
Vanilla 10-20 players4-8 GB
Light modpack (50 mods)4-6 GB
Medium modpack (100-200 mods)6-8 GB
Heavy modpack (300+ mods)8-12 GB
Large community (50+ players)16+ GB

Where you can run it

A network runs in one place, so the region is a decision for the whole thing at once. You make it at checkout: 9 locations across the US, the UK, the EU, and Asia-Pacific are included, and the Germany (EU) region costs nothing on top of the plan price. This plan is also available on 3 extra regions in the wider network behind Astroworld, next to the regions we run on our own hardware.