We estimate the number of users by analyzing the requests induced by clients to relays and bridges.
This graph shows the estimated number of clients connecting via bridges. These numbers are derived from directory requests counted on bridges. Bridges distinguish connecting clients by transport protocol, which may include pluggable transports, so that graphs are available for different transports. For further details see these questions and answers about user statistics.
The following events have been manually collected on this wiki page and might be related to the displayed graph.
|Dates||Places/Protocols||Description and Links|
macOS 10.12 (Sierra) is released, breaking some programs that are built with Go <1.7, including the meek-client that comes with Tor Browser. (See 2016-11-15 unbreaking event.)
geoip and geoip6 databases updated to "September 6 2016 Maxmind GeoLite2 Country" (
Iran intensifies the blocking begun on 2016-08-20, getting most of the remaining direct users. There is interference in the graphs from the bridge authority changeover on 2016-09-02, but because the changeover would not have affected counts of ''direct'' users, it may be a coincidence.
The former bridge authority Tonga shuts down. Bridges that have not updated to tor 0.2.8.7 or 0.2.9.2-alpha (which include all 5 default obfs3 bridges and 3/16 default obfs4 bridges) stop reporting statistics.
shutdown notice blog post loss of reporting bridges
CollecTor begins publishing bridge stats from the new bridge authority Bifroest.
tor 0.2.8.7 and 0.2.9.2-alpha are released, changing the bridge authority from Tonga to Bifroest.
|2016-08-24||China <OR> Unknown||
Large decrease in users in China.
Iran somehow blocks most direct Tor connections. May also affect bridge users, but it's hard to tell because there were few vanilla bridge users anyway.
Changed the meek-amazon CDN price class from "All" to "Use Only US, Canada and Europe."
|2016-08-18 to 2016-08-29||Mali Unknown||
Sudden increase of direct and bridge users in Mali.
relay graph bridge graph
geoip and geoip6 databases updated to "August 2 2016 Maxmind GeoLite2 Country" (
|2016-07-21 to 2017-03-03||Brazil meek Unknown||
Sustained increase in meek users in Brazil. Locals believe that they are not actual users, rather bots or something like that. End date coincides with shutdown of meek-azure before migration. Would pick up again 2017-06-07.
geoip and geoip6 databases updated to "July 6 2016 Maxmind GeoLite2 Country" (
Coup attempt in Turkey. A paper (§ 4.1) reports a decline in Tor users.
paper relay graph
This material is supported in part by the National Science Foundation under Grant No. CNS-0959138. Any opinions, finding, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation. "Tor" and the "Onion Logo" are registered trademarks of The Tor Project, Inc.. Data on this site is freely available under a CC0 no copyright declaration: To the extent possible under law, the Tor Project has waived all copyright and related or neighboring rights in the data. Graphs are licensed under a Creative Commons Attribution 3.0 United States License.