ViceHubFiveM discovery hub
ScriptsServersStreamersViceAIArticlesMerch
Sign inSign in

Navigate

ScriptsServersStreamersViceAIArticlesMerchSign in
ViceHub
ScriptsServersStreamersArticlesShopsFiveM StatsAboutMethodologyEditorial PolicyContactMerchSubmitSign in
QBCore Scripts·ESX Scripts·Free FiveM Scripts·Best FiveM Scripts·FiveM Server List·Live GTA Streams
© 2026 ViceHub · Independent FiveM discovery hub.Privacy · Terms · Not affiliated with Cfx.re, Rockstar Games, or Take-Two Interactive.
All articles →

FiveM Server Requirements: What Actually Matters (and What Hosts Invent)

ByBoris K.·Founder & Editor, ViceHub
Jul 13, 20266 min read

Every FiveM server requirements table you have read was invented by a company selling servers. Cfx.re publishes no hardware spec at all. Here is what actually determines whether your server runs well.

Search for FiveM server requirements and you will find a dozen neat tables: 4 vCPU, 8GB RAM, 100GB NVMe, "recommended for 64 players". They look authoritative. They cite nothing.

Cfx.re does not publish hardware requirements for FXServer. Not a minimum, not a recommendation, nothing. The only system requirements in the official documentation are for the client, and those are just GTA V's own specs. Every server spec table you have read was written by a company that sells servers.

That is not a conspiracy — a host has to put something on the pricing page. But it does mean the numbers you are anchoring on have no authority behind them, and you should stop treating them as if they do. Here is what actually determines whether a FiveM server runs well.

The one architectural fact that explains everything

FXServer runs a few threads. The one that matters is svMain, and here is the fact that should drive every hardware decision you make:

svMain ticks at 20Hz — a 50ms budget — and every single server-side resource ticks on it.

There is no per-resource isolation and no spreading of script load across cores. Your framework, your inventory, your jobs, your dispatch, your custom scripts — they all share one thread's budget, in sequence.

The consequences fall out immediately:

  • Single-core clock speed is what matters for script performance. Cores 5 through 32 are not helping your Lua run faster.
  • Not "makes it slower" — stalls it. A synchronous loop or a slow SQL query freezes every other resource for every player.

Related articles

How Many Players Does a FiveM Server Need to Feel Alive?
Jul 13, 2026

How Many Players Does a FiveM Server Need to Feel Alive?

The median FiveM server peaked at two players. Not two hundred — two. We measured 31,334 of them. Here is what a server realistically needs to feel alive, and why your slot count is fiction.

server datafivem serverserver owners
7 min readRead
Why Your FiveM Server Isn't Showing in the Server List
Jul 13, 2026

Why Your FiveM Server Isn't Showing in the Server List

Your server boots, you can join by IP, and it is nowhere in the browser. There is a specific list of causes, and one of them — an artifact rule Cfx.re enforces — is the one nobody mentions.

fivem serverfxservertroubleshooting
8 min readRead
One blocking script stalls the entire server.
  • Buying a bigger machine does not fix a bad script. This is the mistake that costs owners the most money. If a resource is eating 80ms of a 50ms budget, a faster CPU makes it 60ms. It does not make it correct.
  • (A pedantic but important note, since you will see it stated both ways: FXServer is not literally single-threaded — networking and sync have their own threads. But all resource execution is on the one thread, and that is what people are reaching for when they say it. Describe it accurately and the buying advice becomes obvious.)

    When svMain overruns, you get this in console:

    server thread hitch warning: timer interval of 384 milliseconds

    That is 384ms against a 50ms budget. Every player felt that as a freeze. That message is the single most important line in your server console, and it points at software, not hardware.

    So what should you actually buy?

    Given the above, here is the honest guidance — offered as reasoning, not as a spec table, because I am not going to invent one either.

    Prioritise single-core clock speed. A CPU with a strong per-core score will outperform a many-core server chip with a lower clock for FiveM's script load. This is the one place where spending money reliably helps.

    Do not buy cores you cannot use. They will sit idle while svMain struggles.

    RAM scales with resources and players, and is rarely your bottleneck. More RAM does not make scripts faster. If your server is hitching, RAM is almost certainly not the reason, and upgrading it is the most common way owners spend money without fixing anything.

    Use fast storage. Assets and MLOs stream from disk; SSD or NVMe is worth having, and it is cheap.

    On bandwidth: there is no official figure, and I am not going to make one up. Cfx.re documents no per-player bandwidth requirement. The per-player numbers you have seen quoted are hosting-company estimates. Measure your own server under load — that number is real; theirs is a guess about someone else's server.

    Windows or Linux? Know this before you decide: Cfx.re's own documentation describes the Linux build as a courtesy port, and says that if you experience issues you are more likely to see them fixed if you use Windows. Plenty of serious servers run Linux successfully. But you should make that choice knowing it is officially the second-class target, rather than discovering it while debugging something obscure.

    Slots, OneSync, and the money question

    Two hard rules from the documentation, which affect what you need to buy:

    • sv_maxclients accepts 1 to 2048. Values starting at 32 require onesync set to on or legacy. Values above 64 require onesync on.
    • OneSync is free up to 48 slots. Above that, it requires a paid Cfx.re tier (Element Club Argentum or higher).

    So there is a real cost cliff at 48 slots — and, from our data, almost nobody should be anywhere near it.

    The reality check, from 31,334 real servers

    ViceHub polls every public FiveM server every five minutes. On a single full day, of the 31,334 servers that had at least one player:

    • 1,800 ever exceeded 32 concurrent players.
    • 1,109 exceeded 64.
    • 608 exceeded 128.

    Roughly 94% of servers never went past 32 players in a day. And of the 9,102 servers configured for 33 to 48 slots, only 41 ever reached even half capacity.

    Read that against the hosting pages you have been browsing. You are being sold hardware for a 64-player server, and the overwhelming likelihood — statistically, overwhelmingly — is that you will peak in single digits for a long time. Our full population data shows the median public server peaks at two players.

    This is not discouragement. It is a purchasing argument: buy for the server you have, not the one on your moodboard. You can upgrade the machine in an afternoon on the day you actually need to. You cannot get back the money you spent on a year of capacity you never used.

    What actually causes lag

    When someone says "my server is lagging", they usually mean one of four unrelated things. Diagnose which:

    1. Client frame drops. The player's FPS is low. This is a client-side problem — graphics settings, mods, or an expensive resource drawing every frame. Check with resmon.
    2. Server hitches. Everyone freezes at once. This is svMain overrunning, and it is almost always a script or a database query — not hardware. The console tells you outright.
    3. Network lag. High ping, rubber-banding. Check netgraph. This is routing, hosting location, or bandwidth — genuinely a hosting problem, and the only one of the four that buying a better host reliably fixes.
    4. Streaming stutter. Hitching when moving through the map or entering an MLO. This is asset streaming, and it shows up in resmon's Streaming column, not in CPU ms.

    They have different causes and different fixes, and people spend money on the wrong one constantly. The most expensive mistake in FiveM hosting is upgrading the CPU to fix a server hitch caused by an unindexed database query.

    Before you buy anything

    1. Profile what you have. Our performance guide covers resmon and the built-in profiler — both free, both already installed.
    2. Read your server console. Hitch warnings name the problem.
    3. Check your database. Unindexed queries on a growing player table are the classic cause of a server that got slower over six months while the hardware never changed.
    4. Only then consider hardware, and when you do, buy clock speed rather than core count.
    5. Set a realistic slot count. You do not need 64 slots. Almost nobody does.

    The honest summary: FiveM server performance is a software problem wearing a hardware costume. The hosts selling you spec tables know this, and the spec table is still the easiest thing to sell. Browse the live server list to see what real servers of every size are actually holding — the population history on each one is the same data used throughout this article.

    Frequently asked questions

    What are the official FiveM server requirements?

    There are none. Cfx.re publishes no minimum or recommended hardware specification for FXServer anywhere in its documentation — the only system requirements it publishes are for the client, and those are GTA V's. Every FiveM server requirements table you have seen was written by a hosting company.

    Is single-core CPU speed more important than core count for FiveM?

    For script performance, yes, and there is a concrete reason. FXServer runs several threads, but every server-side resource ticks on the main thread at 20Hz. So all your scripts share one thread's budget, and a fast core beats many slow cores. Buying a 32-core server will not make a badly written script faster.

    How much RAM does a FiveM server need?

    Less than you have been told, and it scales with resources and players rather than being a fixed number. RAM is rarely the thing that makes a FiveM server perform badly — a blocking script on the main thread is. Do not buy RAM to fix a performance problem you have not diagnosed.

    How much bandwidth does a FiveM server need per player?

    There is no official figure, and anyone quoting you a precise per-player number is guessing. Cfx.re publishes no bandwidth specification. Estimate from your own server's actual usage rather than from a hosting company's marketing page.

    Should I run my FiveM server on Windows or Linux?

    Cfx.re describes the Linux build of FXServer as a courtesy port, and states that if you experience issues you are more likely to see them fixed on Windows. Linux works and plenty of large servers use it, but you should know that it is officially the second-class target before you commit.

    Researched and drafted with AI assistance. How ViceHub guides are made

    fivem serverhostingperformancerequirementsvps

    Join the Discord for more FiveM signal

    Get resource discovery notes, shop and server discussion, server-owner advice, and community updates from the ViceHub crew.

    Join the Discord

    Written by

    Boris K.Founder & Editor, ViceHub

    Built the crawler behind ViceHub's server data. Writes the guides here.

    More from Boris →

    On this page

    The one architectural fact that explains everythingSo what should you actually buy?Slots, OneSync, and the money questionThe reality check, from 31,334 real serversWhat actually causes lagBefore you buy anything
    FiveM resourcesAll articles
    How to Make a FiveM Server in 2026 (txAdmin, Start to Finish)
    Jul 13, 2026

    How to Make a FiveM Server in 2026 (txAdmin, Start to Finish)

    Most FiveM setup guides on Google were written for a version of txAdmin that no longer exists. This one is checked against the current docs, and covers the three-month artifact rule that quietly removes servers from the browser.

    fivem servertxadminfxserver
    6 min readRead