Global Bitcoin nodes by city
14194 cities with their respective number of global IPv4/IPv6 Bitcoin nodes as of Wed Dec 31 00:00:00 2025 UTC.
Window size: 90-day
| RANK | CITY | NODES |
|---|---|---|
| 1 | n/a | 55991 (9.17%) |
| 2 | Germany Berlin | 12879 (2.11%) |
| 3 | Germany Munich | 7549 (1.24%) |
| 4 | Germany Hamburg | 6954 (1.14%) |
| 5 | Thailand Bangkok | 5659 (0.93%) |
| 6 | Germany Frankfurt am Main | 5357 (0.88%) |
| 7 | Australia Sydney | 5272 (0.86%) |
| 8 | Switzerland Zurich | 4877 (0.80%) |
| 9 | United States New York | 4637 (0.76%) |
| 10 | United States Miami | 4378 (0.72%) |
| 11 | United States Chicago | 4129 (0.68%) |
| 12 | Germany Cologne | 4062 (0.67%) |
| 13 | The Netherlands Amsterdam | 3897 (0.64%) |
| 14 | Brazil São Paulo | 3857 (0.63%) |
| 15 | United States Los Angeles | 3833 (0.63%) |
| 16 | Canada Toronto | 3735 (0.61%) |
| 17 | Australia Melbourne | 3688 (0.60%) |
| 18 | United States Dallas | 3639 (0.60%) |
| 19 | United States Atlanta | 3499 (0.57%) |
| 20 | Russia Moscow | 3355 (0.55%) |
| 21 | Spain Madrid | 3335 (0.55%) |
| 22 | Austria Vienna | 2959 (0.48%) |
| 23 | Germany Stuttgart | 2948 (0.48%) |
| 24 | France Paris | 2928 (0.48%) |
| 25 | United States Denver | 2862 (0.47%) |
This page reports the estimated size of the Bitcoin peer-to-peer network including both reachable and unreachable nodes, i.e. global nodes. Unlike the low churn rate estimation method for reachable nodes (see the latest snapshot here), the method for this report can only provide a rough estimation and does not filter out potentially spurious nodes that may be gossiped by non-standard/spam/malicious peers.
Bitnodes crawler captures these nodes from the addr messages returned by all the reachable nodes. Each snapshot or data point in this report represents a rolling window. A snapshot with window size of 1 day will include all nodes by IP addresses with timestamps less than 1 day old. The timestamp for a node here refers to the time when its peer last connects to it. If you turn on your Bitcoin node for only a few minutes anytime during the last 24 hours, it will be included in the latest snapshot with a window size of 1 day.
Multiple nodes from the same IP address, but different port numbers are counted as one node in this report. A larger window size may increase the likelihood of the same node being counted more than once due to, e.g., IP lease renewal.
A Bitcoin node may be unreachable for several reasons. It may be configured by the operator to only attempt to make outgoing connections or it may be located behind corporate/ISP firewalls or NAT. A node could also become temporarily unreachable if it has hit its maximum allowed connections or if it is in the process of syncing up to the latest blocks. As it is impossible to connect to an unreachable node directly, we cannot reliably confirm the true existence of an unreachable node, hence the rough estimation.