Showing observation from: 2026-09-23 18:00:02 UTC (3600s)
| Host | API 1 | Run by | Location | Users | Events | % of max 2 | Uptime 3 |
|---|---|---|---|---|---|---|---|
| atproto.africa | Blacksky Algorithms | 🇺🇸 Virginia (OVH) | 264,616 | 3,378,230 | 128.93% | 95.3% | |
| relay.bas.sh | @bas.sh | 🇩🇪 Germany (Hetzner) | 205,242 | 1,477,789 | 100.00% | 100.0% | |
| asia.firehose.network | @sri.xyz | 🇮🇳 Mumbai (OVH) | 205,186 | 1,471,543 | ⚠️ 99.97% | 100.0% | |
| nyc.firehose.stream | @sri.xyz | 🇺🇸 Virginia (OVH) | 205,185 | 1,521,992 | 99.97% | 100.0% | |
| chennai.firehose.stream | @sri.xyz | 🇮🇳 Mumbai (OVH) | 205,180 | 1,521,981 | 99.97% | 100.0% | |
| sfo.firehose.stream | @sri.xyz | 🇺🇸 Oregon (OVH) | 205,161 | 1,521,848 | 99.96% | 100.0% | |
| relay1.eurosky.network | Eurosky | 🇩🇪 Germany (Hetzner) | 204,984 | 1,470,663 | 99.87% | 100.0% | |
| jetstream1.eurosky.network | Eurosky | 🇩🇪 Germany (Hetzner) | 204,975 | 1,535,988 | 99.87% | 100.0% | |
| relay3.fr.hose.cam | @bad-example.com | 🇫🇷 France (OVH) | 204,928 | 1,471,775 | 99.85% | 100.0% | |
| relay1.us-east.bsky.network | Bluesky PBC | 🇺🇸 Virginia (–) | 204,923 | 1,470,314 | 99.84% | 99.4% | |
| jetstream2.fr.hose.cam | @bad-example.com | 🇩🇪 Germany (Hetzner) | 204,923 | 1,520,056 | 99.84% | 100.0% | |
| bsky.network | Bluesky PBC | 🇺🇸 California (–) | 204,921 | 1,470,240 | 99.84% | 100.0% | |
| relay1.us-west.bsky.network | Bluesky PBC | 🇺🇸 California (–) | 204,921 | 1,470,240 | 99.84% | 100.0% | |
| jetstream2.us-west.bsky.network | Bluesky PBC | 🇺🇸 Oregon (OVH) | 204,921 | 1,520,719 | 99.84% | 99.4% | |
| jetstream1.us-west.bsky.network | Bluesky PBC | 🇺🇸 Oregon (OVH) | 204,921 | 1,520,719 | 99.84% | 100.0% | |
| jet.firehose.stream | @sri.xyz | 🇨🇦 Quebec (OVH) | 204,903 | 1,520,637 | 99.83% | 99.4% | |
| relay.fire.hose.cam | @bad-example.com | 🇨🇦 Montreal (ServaRica) | 204,892 | 1,472,424 | 99.83% | 100.0% | |
| jetstream1.us-east.fire.hose.cam | @bad-example.com | 🇺🇸 Virginia (Hetzner) | 204,882 | 1,520,557 | 99.82% | 100.0% | |
| jetstream.us-west.bsky.network | 2 | Bluesky PBC | 🇺🇸 California (–) | 204,758 | 1,517,604 | 99.76% | 100.0% |
| jetstream.us-east.bsky.network | 2 | Bluesky PBC | 🇺🇸 Virginia (–) | 204,751 | 1,517,587 | 99.76% | 99.4% |
| jetstream2.us-east.bsky.network | Bluesky PBC | 🇺🇸 Virginia (OVH) | 204,680 | 1,518,052 | 99.73% | 99.4% | |
| jetstream1.us-east.bsky.network | Bluesky PBC | 🇺🇸 Virginia (OVH) | 204,499 | 1,511,991 | ⚠️ 99.64% | 100.0% | |
| relay.t4tlabs.net | @isla.pet | 🇨🇦 Toronto (Xenyth) | 203,871 | 1,447,136 | 99.33% | 100.0% | |
| relay.waow.tech | @zzstoatzz.io | 🇩🇪 Germany (Hetzner) | 203,463 | 1,448,526 | 99.13% | 100.0% | |
| jetstream.waow.tech | @zzstoatzz.io | 🇩🇪 Germany (Hetzner) | 203,457 | 1,516,490 | 99.13% | 100.0% | |
| zlay.waow.tech | @zzstoatzz.io | 🇩🇪 Germany (Hetzner) | 202,272 | 1,441,086 | 98.55% | 99.4% | |
| europe.firehose.network | @sri.xyz | 🇵🇱 Warsaw (OVH) | 197,346 | 1,422,030 | 96.15% | 100.0% | |
| frankfurt.firehose.stream | @sri.xyz | 🇩🇪 Frankfurt (OVH) | 197,339 | 1,470,694 | 96.15% | 100.0% | |
| london.firehose.stream | @sri.xyz | 🇬🇧 London (OVH) | 197,337 | 1,470,664 | 96.15% | 100.0% | |
| northamerica.firehose.network | @sri.xyz | 🇺🇸 Virginia (OVH) | 165,124 | 1,078,053 | ⚠️ 80.45% | 100.0% | |
| jetstream.fire.hose.cam | @bad-example.com | 🇨🇦 Montreal (ServaRica) | 97,963 | 540,135 | 47.73% | 100.0% | |
| relay.upcloud.world | UpCloud | 🇵🇱 Warsaw (–) | 72,135 | 503,213 | 35.15% | 100.0% | |
| relay.klbr.net | @klbr.net | 🇨🇦 Canada (-) | 2,574 | 4,065 | ⚠️ 1.25% | 12.4% | |
| relay.klbr.net | @klbr.net | 🇨🇦 Canada (-) | 2,407 | 3,674 | ⚠️ 1.17% | 12.4% | |
| relay.feeds.blue | @mackuba.eu | 🇩🇪 Germany (Netcup) | 635 | 6,642 | 0.31% | 100.0% |
1) = firehose (CBOR), = Jetstream, 2 = Jetstream v2
2) Outlier results with suspicious numbers are excluded from calculating the "max users" value, which is why some entries might appear with % larger than 100. "No signal received" means the test runner couldn't make a connection, while 0 events mean that no data was sent over the websocket.
3) Uptime is calculated by making connections to listed servers randomly during the day, around 4 times per day on average for a given source, and recording if a connection was successful and received any data. The % is an average over the last 30 days.