Skip Navigation

Spreading of the 100 biggest Lemmy communities

The biggest surprise for me was the https://hexbear.net count, an instance I hardly interact with.

Community CountCommunity Subscriber Count
beehaw.org6133450
hexbear.net33663204
lemdro.id117052
lemmy.blahaj.zone115907
lemmy.dbzer0.com153006
lemmy.ml14356460
lemmy.one116257
lemmy.world39851950
lemmynsfw.com233586
sh.itjust.works116006
sopuli.xyz114093

The data this is based on comes from https://lemmyverse.net where you can just download a full json of the data they have (I excluded all communities marked as "suspicious")

EDIT: The data if you sort by active users last month:

Community CountCommunity Active Month Count
awful.systems12616
feddit.org27363
feddit.uk25289
hexbear.net12952
lemdro.id12898
lemm.ee38898
lemmy.blahaj.zone111422
lemmy.ca314910
lemmy.dbzer0.com313752
lemmy.ml1054949
lemmy.world57338384
lemmy.wtf13602
lemmy.zip312020
mander.xyz111469
sh.itjust.works537365
slrpnk.net310897
sopuli.xyz210070
ttrpg.network14107

Community Count:

Community Users:

433 comments
  • Surprised I dont see programming.dev in the data, we definitely have at least 3 communities in the top 100 (programmer_humor, programming, linux)

    • Manually counted communities in the top 100 per instance and threw it into another pie chart (for active users / month)

      This also seems to be different than the results gotten from lemmyverse as the lemmyverse data hasnt been updated in 11 days according to that site

      A bunch of instances gained or lost some coms in the top 100 from variance of things happening in the last week

      (the eight instances that it decided to not give labels to that have 1 community are feddit.uk, lemmy.zip, beehaw.org, lemdro.id, ttrpg.network, lemmy.wtf, lemmy.blahaj.zone, mander.xyz)

      edit: updated graph to be more accurate users/month counts

    • In this list it doesn't seem like it: https://lemmyverse.net/communities?order=active_month

      • Seems like lemmyverse doesnt have the instance listed at all for some reason, assuming a crawling issue. I reported it on their repository. Would be new since I remember it showing the instance before

        You can check in https://programming.dev/communities that programmer humor has way more active users than most communities here

  • Probably unintended side-effect of this post: A few people like me discovering new communities to follow. Thank you!

  • A barchart might be better as the comparison of instances with the most subscribed accounts doesn't mean much I feel

    we have some users that register but are inactive and/or are infrequently active which could be a sign of lurkers or bots but empty accounts don't mean much when it comes to the health of an instance.

    However; if we look at each community's active monthly and daily users it can tell another story and that data compared against Reddit's could be useful for anyone seeking alternatives

    I'm rambling with little sleep but hopefully what I've said make a little bit of sense

  • Is there a way to create a pie like this but on an individual basis?

    • You mean the spread of the communities you are subscribed to? If so: probably yes, but not an easy one as you have to have access to the data (or more easily: the database)

433 comments