Run an exit node in ten minutes

Exit capacity is the scarce resource in any VPN. Anyone with a small box and a stable line can supply it, and the client never has to configure a thing.

1. Get a machine

A 1 vCPU instance with 512 MB of RAM and a gigabit port is enough for a few hundred concurrent sessions. The daemon is written to stay under 28 bytes of per-node state, so memory is not the constraint — bandwidth is.

2. Start the daemon

nimwire-node start --wallet "41d7FXjswpK1111111111111111111111111111111111111111111111111111111111111111111111111111112KhNi4"

The wallet address is where payouts land. It is also the identity other nodes use to reach yours, so do not rotate it casually.

3. Add upstreams if you need them

If your provider blocks the usual ports or you want to egress through another network, hand the daemon one or more proxies. Every documented scheme is supported:

nimwire-node start \
  --wallet "41d7FXjswpK1111111111111111111111111111111111111111111111111111111111111111111111111111112KhNi4" \
  --proxy socks5://user:pass@185.22.14.9:1080 \
  --proxy socks4://45.12.88.3:1085 \
  --proxy http://91.108.4.12:8080 \
  --proxy-file ./proxies.txt

Each proxy becomes a virtual exit. Clients see ordinary countries; they never see, and never configure, a proxy.

4. Keep it honest

Nodes that fail the forwarding proof are dropped from rotation. There is no penalty beyond that, and no way to buy your way back in. Relaying correctly is the whole job.

What you earn

Traffic is metered per byte and paid in XMR to the bound address. The dashboard shows per-node earnings, the payout address and the online state; earnings stay in the node address until you sweep them.

Operational notes

• Use HTTP/3 only. There is no HTTP/1.1 or HTTP/2 fallback to misconfigure.
• Terminate TLS at the daemon; do not put a reverse proxy in front of it.
• Watch disk: the daemon writes configuration and nothing else.
• One machine can host many virtual exits. Start small and add capacity when your CPU idles.