Hi,guys,I’m really really really curious how popular haiku is now.
So if you’re using haiku more than 14 hours every week in 2026,please reply, adding which country you are from is much appreiciated
HaikuDepot’s server logs are probably the only useful way to get an approximate count of Haiku users?
The number of clients checking for updates daily/weekly/monthly would under-estimate total users, but would be a decent proxy measure.
1 Like
I’m not considering meself a proper Haiku user yet, but I expect to be one soon. 
The forum has 12500 registered users,but not all Haiku users are in the forum,and not everyone who registered years ago is still using Haiku.
Anyway,it’s as much as you can get for an estimate,and it’s probably enough.
I’m personally very happy that Haiku doesn’t have any form of tracking included and doesn’t make unwanted network requests.
Some other numbers you can base estimates on: The Haiku repository on Github has about 2200 stars and the Haiku User Map has 375 registered users.
1 Like
FWIW my Youtube channel now has 527 subscribers, but non-subscriber views outnumber subscriber views.
Also therer are 177 followers on the Haiku Flipboard magazine.
4 Likes
There is no user-tracking happening in HDS so that’s not possible right now. Maybe at some point it would be nice to add some platform level opt-in telemetry but it would be quite a bit of effort to take through a design doc → implementation → rollout.
In the short term from HDS, you can login then go menu → Reports → General Metrics Reports and run it. This will yield a CSV file (I see the filename is currently wrong; I’ll look into that but for now just rename it to .txt).
The top section “Desktop Versions Traffic” shows the versions of HaikuDepot that are communicating. You can find the current version and history of versions on src/apps/haikudepot/HaikuDepot.rdef in the mono-repo.
There is an older API called V1 for HDS which is still offered for very old versions of HaikuDepot; this is captured under “API Generation”. I may get rid of V1 in late 2026 or 2027.
Multi- / Single-page shows how many users are using the Single Page Application (AngularJS) verses the simple HTML interface (Multi). The multi page option will be counting crawlers so this is the reason why the volumes are much higher.
The Bulk Downloads section shows how often the HaikuDepot desktop application is calling for various sorts of data refresh; this would be a rough idea of how many times the desktop application is started up.
Unfortunately these metrics in HDS are held in memory only and are only counters since HDS was last restarted. For admin users there is an indicator in the UI as to when the system was restarted but regular users can’t see this. These metrics will also stop being truly meaningful when >1 HDS instance is deployed because the stats will only be reported for the instance the report is run in.
To resolve both problems properly would require deploying a metrics system such as Prometheus + Grafana but I don’t have time to deal with that right now and I’m guessing it would be quite a bit of work for the admin team.
Oh, I wasn’t suggesting there was any telemetry in the app; my apologies for being unclear.
I meant the web server logs, where at the very least you can get a simple count of IPs doing daily/weekly/monthly/ever refreshes of the repo.
I know about the user map,but I think it’s outdated
1 Like
I don’t think there’s a quick and easy solution to this.
If we made an application, which people had to install in their system, in order to “count users”, some people would install it, others wouldn’t.
If we counted the number of installations, people like me, who may be reinstalling all the time, would ruin the statistics.
… etc. …
Haiku Insider receive 100 until 500 visitors every day, so it might be another statistics which can be used (figures from the default admin tool provided by Bludit tool)
1 Like
What will it good for?
1 Like
Does 14 hours a day also count? 
4 Likes
Well an average Haiku users, may spend perhaps about 1 or 2 hours a day on Haiku. so it should count 7 times. 
1 Like
I’d rather not count. Low numbers are discouraging, hi numbers invite unwanted influences.
In many ways, the longer we stay under the radar the better
The only important number at this time is the number of active and competent developers we have.
5 Likes
The map is still maintained,but the “where are you from” topic didn’t receive any new posts for some time:
Also,no pull requests to the repository at: haiku-usermap/pages: A (unofficial) map showing where in the world the Haiku developers and users are. - Codeberg.org
2 Likes
more users means more donation, developers need money to live
no ,some users may not have much time using pcs during weekday, they may use pc on weekend more. that’s why I don’t say 2 hours per day
Bean counting does not add more beans 
2 Likes
Indeed, spill the beans, brother.
One might even say that bean counting takes time away from bean farming.
2 Likes
This is not entirely true : most of the developers on haiku develop it in their free time. So money doesn’t make sense at that level.
However if tomorrow there’s a huge gap in the donation, it’s true that it will be useful (architecture, other additional paid contract, etc)
1 Like