IPv4 has about four billion addresses and they are scarce, which is a large part of what you pay for in a proxy. IPv6 has so many that a single customer is routinely handed a /64 block of 2^64 addresses. That makes IPv6 proxies cheap and practically unlimited in number.

The catches

  • Reach. An IPv6 exit can only reach a site that has an IPv6 address. Large platforms and CDN-fronted sites usually do; many smaller sites and APIs still do not, and a request to them fails.
  • Block size. Because one customer gets a whole /64 or more, sites that take IPv6 seriously rate-limit and block by prefix, not by address. A thousand addresses from one /64 can count as one visitor. What matters is how many distinct prefixes the provider gives you.
  • Reputation. IPv6 proxy space is often hosting space, with the same scoring as any datacenter proxy.

How to check

First, does the target have an IPv6 address?

dig AAAA example.com +short

An empty answer means an IPv6-only proxy cannot reach it. Then check which address family your proxy exits on; this echo service answers over both:

curl -s -x http://USERNAME:PASSWORD@HOST:PORT https://api64.ipify.org

An address with colons is IPv6.

Common confusion

Buying IPv6 proxies to beat per-IP limits works only on sites that count per address. Ask any provider which address family you get before you buy, and test your target with a few addresses first. CIDR notation explains the prefix sizes behind the grouping.