Counter-Strike 2 operator on a Mirage-style map with CSbePRO branding

CS2 Community Servers

Real players on CS2 community servers from the Steam server list, checked every 10 minutes, with fake servers and bots filtered out

Every 10 minutes I take the whole Steam server list for CS2 and ask each server itself who is playing on it. The number below is my estimate of real players on those servers: not what the server browser shows, but what the servers answer when asked one by one. Valve's own matchmaking - Premier, Competitive, Casual, Deathmatch - is not in the Steam server list, so it is not here either. This page is about the community side of the game: public servers, surf, bhop, retakes, 1v1 arenas, zombie escape and everything else people host themselves.

The list is dirtier than it looks. About half of the 40 thousand addresses in it are fake: servers with 255 slots, a made-up game version, or one player list copied to hundreds of ports. They are caught and not counted. The chart shows both numbers - what servers report and how many players appear to be real.

Shortcode

Live Updated 21:16 UTC Next check in ~1 min

135,374people on CS2 community servers right now

On 7,902 servers. Every server is asked one by one who is on it: fake servers, bots and hidden player lists are left out. Valve matchmaking is not in the Steam server list, so it is not here.

Servers in the Steam list claimwhat the in-game browser adds up to, bots off, on 42,385 servers 244,430
Fake playerson 24,493 fake servers, nobody is there 94,434
Hidden player listsservers that don't show who plays, so I can't check them 13,470
Legacy CS:GO and silent serversold CS:GO builds (14) and servers that did not answer 1,152
Real playerson 7,902 servers, bots taken off 135,374

Every community server in the Steam list
Hostings with hundreds of servers, each checked on its own
VAC and non-VAC servers
Players and bots told apart
Valve matchmaking: Premier, Competitive, Casual, Deathmatch

Peak today159,146at 14:06 UTC

Record164,224on 26 Oct 2026, counting since 24 Oct 2026

Steam: game open877,610people with CS2 open in Steam, matchmaking and the menu included Busiest mapde_mirage46,174 players on 2,745 servers Most players, by server locationRussia

RU91,207 players on 4,869 servers

Bots8,482on servers with real players, not counted

What is in the Steam server list

every address, now

24,493fake servers, 57.8%
588legacy CS:GO builds
9,402empty, hidden or not answering
7,902with real players, 18.6%

In CS2 the fakes are not after your players, they are after the list: thousands of addresses with 255 slots, a made-up version or one player list copied to hundreds of ports. 20,133 of them are so obvious they are dropped before I even ask them; the rest fall on the checks. Big hostings with hundreds of real servers on one IP are a normal thing here, so an IP with many servers is never a mark against it - every port is asked and judged on its own. How I spot the fakes

Players over time

real vs claimed

Real playersWhat servers claimFake or bots

Loading the history...

The gap between the two lines is what servers claim and nobody plays: fake players, hidden lists and bots.

When CS2 community servers are played

average real players, your time

Game modes

real players now

Every game mode, what it is and who plays it

Maps right now

real players

1de_mirage2,745 servers46,174
2de_dust21,627 servers22,519
3sandstone_fb429 servers16,036
4de_mirage_fps250 servers7,136
5awp_lego_2241 servers5,437
6awp_lego_2_fb140 servers4,860
7ze_voodoo_islands_p109 servers4,305
8de_mirage_christmas100 servers2,896
9surf_ski_3_nova_kg92 servers2,000
10<null>55 servers1,640

Every map, 30-day ranking

Countries

by server location

Every country, 30-day ranking

Servers

busiest first

Data from the Steam master server list and from every server itself, checked every 10 minutes from my own machine. Server location by IP. The number of people with CS2 open comes from CS2 Player Count.

What the Numbers Mean

Every block of the page, top to bottom

Big number: Real players on CS2 community servers from the Steam list at the last check, and on how many servers. Only servers that answered and whose player list checks out are counted. Bots are taken off, fake servers are left out completely.
How it adds up: The column next to the big number reads top to bottom. First, what every server in the Steam list claims, bots off. Minus the players on fake servers. Minus the players on servers that hide their player list, because I can't check them. Minus people on old CS:GO builds and on servers that did not answer. What is left is the big number.
Updates: The servers are checked every 10 minutes, and a full round takes about two and a half. The badges show when the last check finished and when the next one is due. When new numbers come in, the page changes them by itself - you don't need to reload it.
What is counted: The row of marks under the card: every community server in the Steam list, big hostings included, VAC and non-VAC, with players and bots told apart. Valve matchmaking is not in the list and not counted.
Tiles: Peak today (in your time) and the record, Steam's own count of people with CS2 open (see why it is far bigger), the busiest map, the country with the most players by server location, and bots.
What is in the Steam server list: Every address in the list, split four ways: fake servers (red), old CS:GO builds (amber), servers with real players (green) and the rest - empty servers, hidden lists and servers that didn't answer.
Players over time: The history, from the last 24 hours to everything since I started counting. The bright line is real players, the dashed one is what servers claim, the red hatching between them shows player counts excluded by the checks, including fake servers and hidden player lists. Hover the chart for the exact numbers at any moment. A dashed yellow line marks a moment when the way of counting changed, so a jump there is not a jump in online.
When CS2 community servers are played: Every hour of the week, brighter means more people. The times are in your own time zone, and the square for right now is outlined. Under it: the busiest and the quietest hour and how much the weekend adds.
Modes, maps, countries: Where the real players are right now. The mode is worked out from the server name, the tags it sends to the Steam browser and the map: "BHOP MIRAGE" is Bhop, duels_ maps are 1v1, ze_ maps are Zombie, and so on. The country is where the server stands, not where the players are from. Click any mode, map or country to open its own page.
Servers: Every server with real players, busiest first. Search by name or IP, filter by mode and country, or switch the list to hidden lists, fakes, legacy CS:GO builds and servers that did not answer.

How Fake Servers Are Caught

Half of the list goes before the first question is even asked

The Steam list for CS2 holds around 40 thousand addresses, and about half of them are fake. In CS 1.6 the fakes claim millions of players; in CS2 they are after something else - a place in the list. Thousands of addresses with 255 slots, a game version that does not exist, one and the same player list shown from hundreds of ports. Together they claim far fewer players than they take up rows, but every one of them would sit in the server browser next to a real server.

Steam returns at most 10,000 servers per request and cuts the rest off without saying so. So I ask again and again, by region and by filter, each time leaving out the addresses already found, until nothing new turns up. That is how the list grows from the 29 thousand a plain query shows to the 40 thousand I actually check.

Before asking

Some things are visible in the list itself, and such servers are dropped without a single query:

Slots. More than 64 or zero. CS2 can't hold more than 64, and the fakes love 255.
Version. Not the version Steam says is current, and not one that a good share of VAC servers run (that covers the hours after an update). Pure nonsense like 9.99.9.9 is always fake.
Empty fields. No name, no map or no game folder.
Redirect. The name advertises another server's IP.

After asking

Everybody else gets two questions: who are you and who is playing. A fake can lie about the count, but it has a hard time keeping the lie consistent between two checks ten minutes apart:

Check	What gives a fake away	Why a real server passes
List vs count	The server says 6+ players, the list has less than a third of that. The "255/255" servers answer with 0-2 names	A normal server sends everyone who is on it
Mirror	The same player list - the same names with the same scores - comes from three or more addresses at once	One person can't play on three servers at once. Two addresses are allowed, because one server can show up on two ports
Clone	Ten or more addresses on two or more IPs with one name and one map all show the same player count and the same list length, and they are most of the servers under that name	Real servers under one brand have different numbers at any moment. Full servers are not counted, a full server is naturally the same as the next full one
Time on server	Most players' time on server didn't grow by the time between two checks. For servers that hide names, not a single (score, time) pair continued	Players who remain connected have their time on server increase between checks. A real server can also have a complete player turnover. Bots are left out, and right after a map change the check is skipped
Zero time	Three or more players all with zero time on server, empty names or the same score	Real people join one by one. The first 3 minutes of a map don't count
Too many	More players than slots, plus one	One extra is fine: a reserved slot for an admin shows as 65/64
Joined together	Five or more players with the same time on server and zero frags, and they are most of the list	Real players can join at the same time, but identical connection times and zero frags across most of the player list are suspicious
New name every day	The address takes a completely different name three or more times in a week	A match score or a map in the name is not a new name. Small edits don't count

What Each Status Means

Real players - the server answered and the list holds up. Its players are counted, minus bots.
Fake - caught by the list or by at least one check. Nothing from it is counted, and its page says which check caught it.
Legacy CS:GO - the server runs an old CS:GO build, not CS2. Steam keeps both under one app, so they share the list. Shown apart from the CS2 number, not added to it.
Hidden list - the server answers but doesn't say who is playing. Its players are shown as unverified, outside the main number - can't check them, won't count them.
Not answering - the server is in the Steam list but didn't answer. Down, restarting or blocking queries.

No check is perfect, and a fake that copies a real server's behaviour well can slip through for a while. The checks are tuned on live data and will keep changing, but the rule stays the same: never punish an honest server. If your server is marked fake by mistake, see Your Server.

Multiple Servers on One IP Aren't Necessarily Fake

Hundreds of servers on one IP is the normal thing in CS2

In CS 1.6 an IP with hundreds of servers is almost always a fake farm. In CS2 it is almost always a hosting: Cybershoke runs five hundred match servers from one address, FB-CS.RU fills dozens of 64-slot publics on a handful of IPs, XPLAY spins up a server for every match. Those servers answer honestly - 63 of 64 players, 63 different names, time on server growing between checks - and together they hold most of the real players on this page.

So the number of servers on an IP is never a mark against a server here. Every port is asked and judged on its own, and servers run by hosting providers keep their history like any other. A server page says how many servers share its IP, for information only.

The tool for CS 1.6 works the other way round for a reason: CS 1.6 Steam Server Stats explains what a fake farm is and why it is treated as one there.

Server, Map, Mode and Country Pages

Every server, map, game mode and country has its own history

Click a server in the list and you get its page: real players now, the average and the peak over 7 and 30 days, the all-time record, how often it was up, its chart and its own "when people play" grid, the maps it runs and similar servers in the same country. Play joins it through Steam, and the connect line is there to paste into the console.

Maps work the same way. The map list ranks every map by real players over 7 days, 30 days or a year, with the trend against the period before. A map page shows how many people play it right now, its history and the servers running it at this moment. Don't expect the Premier pool here: on community servers Mirage alone holds more people than every other map together, and awp_lego_2 beats Nuke by a mile.

Game modes and countries have pages too. The mode list explains what Retake, Surf, Bhop, 1v1 or Zombie Escape actually are and how many people play each of them, and the country list shows where the servers stand. Each page has its chart, its busiest hours, its best servers and the maps played there right now.

Before you pick a server to play on, look at its page. An average of 40 players over the week says more than a "64/64" in the browser.

Why Steam Shows a Bigger Number

Two different things are being counted

Steam counts everyone who has CS2 running. This page counts people I can see on community servers from the Steam list who pass the checks. Steam's number is several times higher because it also includes players outside the community servers counted here.

Steam counts, I don't: people in Premier, Competitive, Casual and Deathmatch matchmaking, in the main menu, in the inventory, on servers that hide their player list and on servers that are not in the Steam list.
I count, Steam doesn't: nothing extra. Every player on a community server is also in Steam's number, because CS2 only runs through Steam.

So this page is a part of Steam's number, the part that happens on community servers. The whole number, every 5 minutes with charts and peaks, is on CS2 Player Count.

Your Server on This Page

Adding it, fixing it, and what to show your players

It is on Steam's list. Then it's already here - every server in the master list is picked up within 10 minutes. Find it in the server list by name or IP.
It is not on Steam's list. Then it is not on this page, because the page covers the Steam list only. Check that the server has a game server login token (GSLT), that it is not set to hidden, and that its port is open to the world.
It is marked fake and shouldn't be. Write on the forum with the IP. Its page shows which check caught it, which usually tells what's wrong - a plugin that rewrites the player list, a mirror of the server on several addresses, or slots set above 64.
It hides its player list. Then its players can't be counted as real. Let the server answer player queries. If it passes the other checks, it moves to "Real players" by the next check.

A server page with a week of honest numbers is a good thing to link from your own site or Discord - "40 people on average every evening" is something a new player believes.

Data Coverage and Limitations

What this page can and can't tell

Only the Steam list. Valve matchmaking is not in it, and neither are servers that stay off the list on purpose.
Every 10 minutes. Someone who joins and leaves between two checks is not counted. Over longer periods, the averages become more representative, but short sessions between checks can still be missed.
Country of the server, not the player. A server in Germany full of players from Russia counts as Germany.
Modes are guessed from the server name, its tags and the map. A public server that runs one surf map now and then may switch mode for that hour.
No player names. They are used inside the checks and thrown away. This page only ever shows numbers.
Some honest servers look hidden. A real server that hides its player list with a plugin ends up in "unverified", not in the big number.
Some big servers come and go. A few hostings show a full server at one check and an empty one at the next. The checks treat that as it comes; the averages smooth it out.
History starts on 24 October 2026. The 7-day and 30-day averages fill in over the first weeks, and the "All" range grows from there.

CS2 Community Servers FAQ

Short answers

A glowing FAQ sign on a futuristic holographic panel surrounded by floating question marks

How many people play on CS2 community servers right now?

The big number at the top of this page, updated every 10 minutes. It counts real people on servers from the Steam server list, with fake servers and bots filtered out.

Why is it so much lower than the CS2 player count?

Because Steam counts everyone with CS2 running, while this page only counts players on community servers. This page covers community servers only. The full number is on CS2 Player Count.

Why is it lower than the server browser shows?

About half of the servers in the browser are fake: 255 slots, a made-up version, a player list copied to hundreds of ports. Compare the two lines on the chart - the hatched gap shows excluded player counts, including fake servers and hidden player lists.

Is a hosting with 500 servers on one IP a fake?

No. In CS2 that is how the big platforms work, and their servers answer honestly. Every port is checked on its own. See Hostings Are Not Fakes.

What is "Legacy CS:GO"?

Servers that run the old CS:GO build. Steam keeps CS:GO and CS2 under one app id, so they share the server list. They are shown with their own status and are not added to the CS2 number.

When is the best time to find a full server?

Look at "When CS2 community servers are played". It's in your time zone, and the brightest squares are the hours with the most people. For a single server or map, its own page has the same grid.

What is the most played CS2 map on community servers?

Mirage, by a huge margin, then Dust2. See the map list, ranked by real players over 7 days, 30 days or a year.

How do you know a server is fake?

By the list first: slots over 64, a version that does not exist, empty fields. Then by asking: a list that doesn't match the count, the same player list on several addresses, time on server that doesn't grow, everyone joining in the same second. Details in How Fakes Are Caught.

My server is marked fake. What now?

Open its page - it says which check caught it. If it's a mistake, post the IP on the forum.

Do you store who plays where?

No. Player names are used for the fake checks and thrown away. Only numbers are kept.

Is there the same thing for CS 1.6?

Yes, CS 1.6 Steam Server Stats. The checks there are tuned for 1.6, where the fakes behave differently.

More Steam Tools

The other tools on this site that work with Steam: accounts, demos, inventories and live server numbers

CS2 Community Servers

What is in the Steam server list

Players over time

When CS2 community servers are played

Game modes

Maps right now

Countries

Servers

What the Numbers Mean

How Fake Servers Are Caught

Before asking

After asking

What Each Status Means

Multiple Servers on One IP Aren't Necessarily Fake

Server, Map, Mode and Country Pages

Why Steam Shows a Bigger Number

Your Server on This Page

Data Coverage and Limitations

CS2 Community Servers FAQ

More Steam Tools

Leave a Reply Cancel reply

CS 1.6

Forum

CS2

Pro scene

Interactive

CS 1.6 Tools

CS2 Tools

Steam Tools

CS2 Skins

CS2 Console Commands