I've been curious about something lately.
Nostr has millions of public keys, plenty of activity, and feeds that sometimes seem incredibly busy.
But how much of that activity is actually human?
I went looking for numbers, and here's what I found.
- 67.1% human/non-bot accounts
- 32.9% bots or automated feeds
Those numbers come from a February 2026 peer-reviewed study examining 73 authors and 225 posts.
**Important:** That's a small, specialized sample, NOT a measurement of all Nostr.
Other findings are interesting:
- 20,000+ active humans were estimated on Primal alone in March 2026.
- 50,000+ humans across Nostr has been suggested, but remains speculative.
- A June 2026 independent analysis estimated just **1–2% first-week retention** among newly observed public keys. Those keys weren't all verified humans.
And here's the problem.
One human can have several keys. One bot can generate thousands. Automated reposting can make a network look considerably busier than it really is.
The conclusion?
Nobody currently has a reliable, independently audited percentage of real humans versus bots across Nostr.
Nostr absolutely has a real human community. The question is how large that community actually is, and whether it's growing.
I'd personally rather see 30,000 genuinely active humans than 3 million automated accounts pretending to be a thriving network.
What do you think? Are we measuring Nostr's growth, or just its activity?
Research worth reading:
https://daniella.io/nostr-retention-structural-diagnosis/
#nostr #nostrich #plebchain #bitcoin #bots #deadinternettheory #decentralization
