FiveM slot counts versus the players who actually turn up

The number after the slash is the least informative figure on a FiveM listing. In the week to 28 September 2026, 145 servers on JoinGG declared 2,048 slots, the median server's busiest reading was 38 players, and only 37 servers filled up even once. The declared maximum is a setting with a commercial ceiling. How many people turn up is a separate matter.

What a week of readings shows

JoinGG queries every FiveM server it lists every 15 minutes, through the HTTP info and players endpoints the server publishes itself, and records the result once an hour. These figures cover 21 September 10:00 UTC to 28 September 10:00 UTC. There were 1,854 FiveM listings, and 1,598 of them answered at least once; a server's weekly peak is the highest player count among its readings. The methodology page has the full method.

The median weekly peak was 38. A quarter of servers reached 107 at some point in the week, and one in ten reached 249. Only the top 1% got to 1,099. Measured against each server's own declared maximum, the median server's peak came to 26% of it, and 37 servers hit their maximum at least once.

Declared maximums sit far higher. Of the 1,854 listings, 532 declare between 65 and 128 slots and 351 between 49 and 64. Another 319 declare between 257 and 1,023, and 255 declare 1,024 or more, while only 93 declare 32 or fewer.

Where numbers like 2,048 come from

The declared maximum is sv_maxclients in the server's server.cfg. Cfx.re's documentation gives it a range of 1 to 2,048 and ties the upper part of that range to OneSync, FiveM's own sync system: from 32 upward the server needs onesync set to on or legacy, and above 64 it needs on. With OneSync off, the server is a relay and clients fall back on GTA's peer-to-peer networking. With it on, the server keeps track of the world and decides what each client is sent, leaving out players and entities beyond a zone around each player. That is what makes four-figure slot counts workable in the first place, and the OneSync documentation explains the culling in more detail.

Below that hard limit, the ceiling is commercial. Cfx.re's hosting page allows 48 players without a subscription, and 64, 128 and 2,048 on its three Element Club tiers, Argentum, Aurum and Platinum, the last two with OneSync Infinity. Those are exactly the four most common declared maximums in the data: 128 on 365 servers, 64 on 316, 2,048 on 145 and 48 on 91. Forty-eight is also what the example server.cfg in Cfx.re's own setup guide uses, which goes some way to explaining how many servers still carry it.

Big declarations, modest crowds

Of the servers declaring 1,024 slots or more, 231 answered during the week. Their median weekly peak was 134 players, and 80 of them passed 256 at some point: big servers by any standard, running at a fraction of what their listings advertise. Twenty servers recorded a reading of 1,000 players or more, against 255 listings declaring at least 1,024.

The servers sitting at the free ceiling show the same gap at a smaller scale. Of those declaring exactly 48, 74 answered. Their median weekly peak was 10, and one of them reached 48.

None of this makes FiveM quiet. Averaged across the week, the median FiveM server had 9.30 players on it, the highest of the ten games in JoinGG's comparison. Next came Counter-Strike: Source at 3.92 and Counter-Strike 1.6 at 3.02, and both of those count bots, since Valve's query protocol includes them in the player total. The median Minecraft server averaged 0.21. Only 2.1% of the FiveM servers that answered never had a single player all week, against 28.5% for Minecraft. The likely reason is what FiveM servers are for: roleplay communities that people join on purpose, often through an application, and come back to on a schedule.

The recorded peak is a floor

One caution applies to every peak above. For most FiveM servers, JoinGG's readings that week did not land in every hour: only 15 of the 1,854 had a reading in every hour of the day. A server whose busiest hour fell in a gap shows a lower peak than it really reached, so treat each figure as the least the server managed rather than the most.

Choosing the number if you run a server

Set sv_maxclients to the crowd you can host and moderate, within your tier, and raise it when you reach it. A listing that reads 40/2048 looks deserted to someone scrolling past, even when forty people is a busy roleplay evening, while 40/64 looks like a server worth trying. The walk through the FiveM server.cfg, line by line shows where the setting sits among the rest. More slots than your scripts and database have been tested with is a performance problem you have volunteered for, and the general case for buying slots for the crowd you actually have applies to FiveM as much as to anything. A queue resource also changes what full means on FiveM; what a full server actually means separates the count from the queue.

Reading a FiveM listing with this in mind

Compare the players online now with the busiest reading of the past week, and leave the declared maximum until last. A server that sits at 60 most evenings is a busy server whatever its slash says, and one declaring 2,048 with 12 on is a small server with a large number in its config. The live figures for the whole game are on JoinGG's FiveM stats page, and what a player count on a list really tells you covers the traps that apply to every game. For the things no count can show about a roleplay community, choosing a FiveM roleplay server and the broader guide to FiveM servers pick up where this leaves off.