Europa is a new decentralized VPN marketplace built on Nostr and paid exclusively with Bitcoin. It launched this week from @calle, the developer behind the Cashu ecash protocol. The concept is simple but the implications are significant. Instead of trusting a single VPN company with a global server fleet, one privacy policy, and one monthly subscription, Europa lets independent VPN operators list their servers on an open directory. Each operator runs their own infrastructure in their own jurisdiction, writes their own no-log policy, and sets their own price. Users browse the directory, pick an operator, and pay them directly over Lightning or Cashu ecash. The marketplace never touches the money and takes no cut. After payment, the operator delivers a standard WireGuard or OpenVPN config file. No proprietary app, no custom client. Users import it into the same open-source VPN software they would use anywhere else and connect. When the bundle runs out, the tunnel goes quiet. No auto-renewal, no card on file, no surprise charges. The reputation system runs on Nostr. Instead of editorial ratings or paid placements, recommendations and complaints come from people in your social graph. Sign in with a Nostr identity and the directory weights endorsements from accounts you already follow. If Nostr is not your thing, ignore the feature entirely and start with the smallest bundle from an unfamiliar operator to limit your risk. The entire protocol is built on public Nostr events. Operator listings, reputation data, and payment coordination all live on the Nostr network. The Europa website is just one viewer of that data, not the source. Other directories can exist. If Europa makes bad editorial choices, users can leave without losing access to the underlying marketplace. The protocol spec is CC0 (public domain) and the code is MIT licensed. What makes this meaningful is the architecture. Traditional VPN providers consolidate trust into one company in one jurisdiction. If that company gets acquired, served with a warrant, or breached, every customer is exposed at once. Europa distributes that trust across independent operators. Users can spread their traffic across multiple operators they actually chose rather than trusting a single brand to make every decision for them. Anyone with a server can become an operator. Anyone with a Lightning wallet can become a customer. No accounts required for either side. image

Replies (22)

fade2's avatar
fade2 yesterday
Calle how do you accomplish more in two years than I do in twenty?
Also also, the wireguard import via qr doesn't work but that could be on their end bc my native camera picks up the info.
Default avatar
Roboto yesterday
Read this, bullish as fuck!
my friend, @ps in search of a good free VPN because he trying perseverently to befriend #Yggdrasil 🏝️ , #Mycelium 🍄 , nostr, #bittorent / #DHT and so on. Hope it'll helps to him!
That's for sellers, but not users. What I'm trying to figure out is if I just put this directly on my router and 2 cell phones if there is any decrease in speed. I'm assuming yes at some point but don't know how much testing you've done. Or if you're implying that it solely depends on the provider you're using and how many other users they have at that given moment. Idk how bandwidth is managed if say a seller has 10 clients and each of those have 5 devices running at the same time.
1. The operator's pipe ÷ everyone currently using it. Biggest factor — same as any VPN. The operator's listing publishes a capacity field (e.g. "1 Gbps shared, US-East") and that pipe gets split across whoever's actively transferring. The reference daemon doesn't rate-limit per peer; fairness is emergent through TCP backpressure, the same way the rest of the internet handles shared links. In practice most peers are idle most of the time (nobody streams 24/7), so a well-provisioned operator usually delivers near line-rate even with hundreds of peers connected. If an operator over-sells, people switch operators — that's the market-enforced fairness. 2. Your end of the tunnel. WireGuard is fast but not free on your hardware. Modern phones/laptops with hardware crypto easily do several hundred Mbps. A router-as-VPN-client depends entirely on the router: an OpenWRT box on decent hardware (Belkin RT3200, Mikrotik, Ubiquiti) does 300–1000+ Mbps; a cheap consumer router doing software-only crypto might cap at 50–100 Mbps regardless of what the operator gives you. 3. Whether you share or split the config. If you put one WireGuard config on your router, every device behind it shares one "peer" slot at the operator. Router + 2 phones on a single router-level config = 1 peer to the operator, and your router does all the crypto. If instead each phone has its own config, that's 3 peers (router + 2 phones) and counts against the operator's IP pool three times. Realistic expectation: with a well-provisioned operator and decent hardware on your end, expect 70–95% of your underlying line rate for normal browsing. The marketplace win vs traditional VPNs isn't "no slowdown ever" — it's that switching operators is one click if one disappoints, instead of a refund cycle. For the "10 clients × 5 devices" scenario: the operator sees 10 peers, not 50. The 50 devices are invisible to the operator — they all flow through their respective peers' tunnels. If the pipe is 1 Gbps and 7 peers are idle at a given moment, the 3 active peers share 1 Gbps three ways. No protocol-level guarantee, just supply and demand.
Awesome explanation, thank you. I wonder how long before some providers figure out how to pair up and offer any easy switching service with a single point of payment.
There's also a bytes volume configuration as well for the Europa Node. 10 Gb of data allowed for a connection -- for example. Should fix that issue if it were to arise.