nos.lol
nos.lol is a widely used public relay that presents itself plainly: it accepts general notes while trying to avoid spam. That makes it a useful relay to study because its value is not branding, but policy, reach and quiet operational discipline.
Why this relay matters
nos.lol belongs in the relay map because it is one of the endpoints people recognize from default lists and setup advice. It is not a protocol foundation and it is not a personal archive. It is a public relay that tries to keep the commons usable by accepting normal notes while pushing back on spam.
A relay page is useful only when it explains the job the endpoint is doing. nos.lol is not just the string wss://nos.lol. It is an operated place where signed events may be accepted, served, filtered, rate-limited, searched or ignored. That makes it part of the user's trust path, even when the relay never takes custody of a key or a coin.
The practical question is simple: what happens if a client depends on this relay and the relay behaves differently than expected? A note may fail to publish, a reply may not be found, a profile may look empty, a wallet-connect message may miss the listener, or a user may assume censorship where the real issue is relay selection. Relay literacy starts there.
What the metadata says
The current NIP-11 response was reachable during the June 18, 2026 check. It identifies the relay as nos.lol, advertises NIP-01, NIP-02, NIP-04, NIP-09, NIP-11, NIP-28, NIP-40, NIP-45, NIP-70, NIP-77, and gives readers a concrete way to compare the endpoint with other relays.
The relay advertises its software as strfry. Its contact field is not advertised, and the public operator key shown in metadata is c5fadeb5d90d...87147cd9. These fields are not decoration. They give client developers and power users a place to start when a relay breaks, changes policy or becomes important enough to audit.
The advertised limits are also part of the story. Maximum result limit is 500, maximum subscriptions are 20, and maximum message length is 131072. A client that ignores those boundaries may look broken even when the relay is doing exactly what it said it would do.
Policy, limits and trust
The metadata does not advertise payment-required or auth-required access in the current response. That makes the relay easier to test, but it does not make it ownerless, limitless or guaranteed to preserve every event.
The main tension is moderation without ceremony. A relay that filters spam is doing useful work, but any filter also creates edge cases, false positives and an operator-shaped view of what belongs on that endpoint.
Good relay choice is therefore not a popularity contest. The better pattern is to ask what the relay is optimized for: broad public reach, paid access, spam resistance, search, profile discovery, wallet traffic, local community use or long-term archival storage. When the job is clear, the relay becomes easier to trust for that job and easier to replace for everything else.
Where it fits in a relay set
Use nos.lol when a client needs a broad read/write public relay and the user wants a pragmatic default. Pair it with a personal relay or an outbox-aware setup if the content needs a durable home.
A healthy Nostr setup usually mixes roles. One or two broad public relays help with reach. A personal or paid relay can improve durability. A search or directory relay helps discovery. A specialized relay handles wallet, group, media or filter behavior. nos.lol belongs in that map as public anti-spam relay, not as a magic relay that solves every routing problem.
That is why this page links both the live metadata and the surrounding sources. The reader should be able to open the endpoint, compare it with NIP-11, inspect the software or project behind it, and then decide whether the relay belongs in their client, their app defaults or their operational playbook.
How to evaluate it today
Start with the relay information document, then test from a real client. Check whether publishing succeeds, whether reads return expected events, whether filters behave as expected, whether the relay appears in other people's relay lists, and whether a second client can find what the first one wrote.
Then look at failure behavior. Does the relay return clear notices? Does it rate limit loudly or silently? Does authentication fail in a way a normal user can understand? Does the operator publish contact information? Does the software version suggest active maintenance? A relay that answers these questions is easier to recommend than a relay that merely connects.
Finally, never let one endpoint become the whole plan. Nostr works because users can move across clients and relays. The moment a single relay becomes invisible infrastructure for a profile, product or payment flow, it deserves the same scrutiny as any other dependency.