Data reportData reports
How many players do Minecraft servers have?
Measured on every running server we list: what a typical server averages, how much the biggest few hold, why slot counts mislead, and the split by mode.
Most Minecraft servers on a public list have only a handful of players, and a few big servers hold most of the crowd. Over the last 30 days the median server we measured averaged 3.5 players while it was up, and the five biggest servers held 63% of all players on the list.
Data from our own pinger over the last 30 days, with hours of the day shown in your time zone. Every figure on this page is live and recalculated from our latest pings.
Searches for this question usually return the size of Minecraft as a whole: how many people play it every month. Those figures count single player, Realms and friends' worlds as well, and they say nothing about what you will find when you join one server. This report measures that instead.
Headline numbers
Right now 8,377 players are online across the servers we list. Over the last 30 days the median server averaged 3.5 players, and the five biggest held 63% of everyone.
Measured 19 Oct 2026, 22:55 UTCHow we measure
| Measure | Last 30 days |
|---|---|
| Servers playable at least once | 129 |
| Median average players | 3.5 |
| Averaged under 1 player | 56 |
| Averaged 20 players or more | 40 |
| Averaged 100 players or more | 23 |
| Share of all players on the top 5 servers | 63% |
| Share on the top 10 servers | 74% |
| Players across the list in an average hour | 11,960 |
| In the busiest hour | 18,099 at |
| In the quietest hour | 6,893 at |
"Average players" here means the average of a server's player count across every ping while it was answering. Servers that were never playable in the period are left out of this report. They are the subject of the uptime report.
The long tail
- 50
- 51Under 1
- 131 to 5
- 205 to 20
- 1720 to 100
- 18100 to 500
- 5500 or more
- Median server
- 3.5
- Five biggest
- 63%
- Ten biggest
- 74%
Each column counts servers by their average players. Half of the 129 servers average 3.5 players or fewer.
Measured 19 Oct 2026, 22:55 UTCHow we measure
An average under one player is easy to misread. It does not mean one lonely player is always there. It means the server is empty most of the time and someone drops in now and then. Over the last 30 days 56 of the 129 running servers were in that group.
The middle of the chart is where it gets interesting for players: servers big enough that someone is always around, small enough that people learn your name. Those are the servers a ranking by size buries, and the ones how to choose a Minecraft server is about.
A few servers carry the list
Over the last 30 days the biggest single server held 38% of all players on the list.
| Servers, biggest first | Share of all players, last 30 days |
|---|---|
| Top 1 | 38% |
| Top 5 | 63% |
| Top 10 | 74% |
| Top 20 | 88% |
| The smaller half of the list | 0.2% |
The top of this table depends on what a handful of servers report. One large network that reports its whole player base on every address, or a count that is inflated, moves these shares on its own. We exclude the clear cases we find, see how we measured below.
One case we could not settle sits at the top right now. As of 19 October 2026, the biggest listing by 30-day average reported a daily average of 4,000 to 10,500 players until 8 October 2026, then 630 to 720 from 9 October on. A real crowd does not shrink like that overnight, and we could not verify the earlier count. It stays in the 30-day figures until early November 2026, so until then read the top shares and the 30-day network totals on this page as too high.
This is not a verdict on small servers. It does explain why the same few names show up at the top of every list sorted by players, and why a new server competes for attention with servers that have a hundred times its crowd.
Peak and typical
A player count you see on a list is a snapshot, and it depends on the clock. For servers averaging five or more players, the busiest hour of the day had, at the median, 2.7 times as many players as the quietest hour over the last 30 days. So a server showing a big crowd at its peak may average far fewer.
Busiest at with about 18,099 players, quietest at . The busiest hour holds 2.6 times the players of the quietest.
Measured 19 Oct 2026, 22:55 UTCHow we measure
The most common peak hour for those servers was over the last 28 days, while the list as a whole peaked at . Every server page on serverlist.net shows the curve for that server, which is a better guide than one number. The network-wide pattern by hour and weekday is in when Minecraft servers are busiest.
Max player slots are not capacity
The "max" in a player count such as 12/500 is a number the owner typed into a config file. It is not how many players the server can hold.
Over the last 30 days, among the servers that averaged at least one player, the median advertised maximum was 500 slots, and the median server's busiest moment of the month reached 9% of it. Very few servers ever come close to their advertised maximum.
The 500 has a mundane source. Velocity, the proxy most networks on our list run in front of their game servers, ships with show-max-players = 500 in its velocity.toml, and the proxy answers the ping, so that default ends up in the server row. The comment above that line in the default config says it plainly: Velocity does not cap the number of players online, so the 500 is only for display. A server without a proxy reports whatever max-players says in server.properties, which defaults to 20. What a server can really hold depends on its hardware, its plugins and how it is split across backends, none of which the ping reveals.
What busy means in each mode
"Busy" depends on the mode. A minigame lobby needs dozens of players to fill queues, while a whitelisted SMP of 15 can feel lively. Here is the typical running server in each larger mode, and where the top quarter begins, sorted by the median.
A server tagged with three modes counts in all three, so the rows overlap and do not add up to the network total.
Measured 19 Oct 2026, 22:55 UTCHow we measure
Servers can carry several mode tags, so they appear in several rows, and with fewer than ten servers in a row a single server can move the median. Read the small rows as a sketch.
Where the median is high, the typical running server has a real crowd, which suggests players in that mode gather on fewer servers instead of spreading thin. Where the median is low and the top quarter starts much higher, a few busy servers sit above a long tail of quiet ones. SMP is the clearest case: it is the biggest tag on the list, the typical SMP averages 5.0 players, and the top quarter starts at 35.0. Mini-games, the mode that needs a crowd most, has 14 running servers with a median of 0.4. The full ledger with votes is in Minecraft game modes by the numbers.
How many players you need (for owners)
There is no magic number, but player counts show where a server stops feeling empty.
Read it from a new player's side. On a server averaging two players, someone is often there, but a newcomer rarely finds a group to play with. Around five average players, a few people are on for much of the day. Somewhere above twenty, a server can keep itself going: new players find company, so they stay, so the next new player finds company too. Our survival report points the same way: in the six weeks to 17 October 2026, 12 of 53 servers averaging fewer than two players stopped, against 2 of 47 averaging ten or more. Every server page shows its own player curve by hour, so you can see where your own server sits.
What this means in practice:
- Aim your launch, events and first promotion at the hours your audience plays, so the people you attract meet each other instead of arriving one at a time into an empty spawn.
- Advertise slots honestly. A 12/500 row tells a player the server is nearly empty, while 12/40 reads as a small community with room. Behind Velocity, the number to change is
show-max-players. - Do not inflate the count. A real crowd builds and fades over days, so a count that jumps or collapses overnight stands out from outside, and we leave counts we cannot trust out of our figures (see below).
Getting from 2 to 20 is the hard part, and how to get your first 100 players is our guide to it.
How we measured
- Pings. Every public listing is pinged with the Java status ping, and crossplay listings also on their Bedrock port, every minute. Before the evening of 17 October 2026 the interval was about 90 seconds.
- Player count. The headline count of a ping is the larger of the Java and Bedrock answers, never their sum, because a server running Geyser reports the same shared count on both.
- Average players. The mean player count over all pings while the server was answering, from our daily rollups for the last 30 days. Only servers that were playable at least once in that window are included.
- Network totals. The sum of all servers' hourly averages for each hour, then averaged by hour of day.
- Exclusions. Listings that sit on a shared hosting lobby whose count belongs to the whole lobby, or that point at another network's address, are left out of player figures: 4 of 216 public listings right now. Every other listing counts on its own, so a network listed twice is counted twice.
- Blind spots. We count what the server reports. A network that reports its whole player base on every address looks bigger than any one of its servers, and a count can be inflated on purpose. We exclude the clear cases we find.
The ping method, the playable rule and the exclusion counts are documented on the methodology page.
Read next
Keep going.
How to get your first 100 players on a Minecraft server
What our pinger shows about new Minecraft servers, and how to win your first 100 players: retention first, then lists, votes, Discord and seasons.
Most popular Minecraft server game modes, by the numbers
Servers, players and votes for each Minecraft server game mode over the last 30 days, which modes are crowded or open, and SMP servers up close.
