How to get a cheap /32 IPv4 address

A cheap /32 IPv4 address is one public address that a person, service, or customer can actually reach. That distinction matters. Someone hosting a small application, a game server, a VPN endpoint, or a device that must sit on an allowlist does not need a subnet. They need one stable public entry point, delivered to the machine that already runs the work.

The smallest number on an IPv4 price page is not always the cheapest result. An address is useful only when traffic sent to it can reach the intended server. The practical comparison includes the address, the route that delivers it, and any server that must be rented solely to hold it. For a one-service deployment, avoiding an unnecessary move or oversized allocation is usually the largest saving.

sequenceDiagram
  participant Server as Your server
  participant Tunnel as WireGuard tunnel
  participant Taipan as Taipan
  participant Public as Public /32
  participant Visitor as Visitor

  Server->>Tunnel: Open tunnel
  Tunnel->>Taipan: Connect
  Taipan->>Public: Route address
  Public->>Visitor: Reachable

A /32 is one address, and one can be enough

IPv4 prefixes describe how many addresses are included. A /32 is exactly one individual address. It is the normal starting size for one server, one router, one firewall identity, or one application that needs to be reachable directly. Renting one address keeps the cost proportional to the job instead of paying for a block that will remain unused.

One IPv4 does not limit a deployment to one website. A reverse proxy can receive HTTPS traffic for several domain names on the same address and send each request to the correct application. A second address is useful when there is a concrete need for a separate public identity, such as a distinct non-HTTP service, a customer isolation boundary, or a partner firewall that must allow one service but not another. Domain names alone are not a reason to order additional IPv4 addresses.

It is equally important to separate a public address from a local static address. Reserving 192.168.1.20 for a server on a router is good local housekeeping, but it only makes that address predictable inside the local network. It does not create a route from the Internet. The same caution applies to many proxy and dedicated VPN products: they can provide a stable address for outbound requests without providing a route for arbitrary inbound traffic to your own server.

Before paying, ask one direct question: can a client on another Internet connection reach my service through this address? The answer should cover the intended protocol and destination, not just state that the IP is “dedicated.” IANA maintains the registry for IPv4 ranges with special uses; those ranges are not replacements for a normal globally reachable public address. IANA IPv4 Special-Purpose Address Registry

Compare the complete monthly cost

An IPv4 always needs a place to live. When an application is already on a VPS, adding an address from that host is often the simplest choice. The address and service are already at the same network edge, so no additional delivery path is needed. The tradeoff is that the VPS remains part of the bill and the workload remains tied to that provider.

That calculation changes when the server already exists elsewhere. A home server may hold local data and storage. An office server may need access to equipment on the LAN. A colocated machine may have more useful hardware than a small VPS. Moving any of those systems merely to obtain a public IPv4 creates migration work, extra storage costs, and another service to maintain. A routed address is designed for this case: the public IP is held at the provider edge and delivered to the server you keep.

Taipan Transit currently lists tunnel access from €2 per month and routed IPv4 addresses from €0.70 per month on its pricing page. For one address, the displayed starting total is €2.70 per month before applicable taxes and any later change to the offer. This is meaningful when the customer already has a suitable server and needs only one public route. Check the live pricing page before ordering; an article should never be treated as a permanent price list.

Traffic also belongs in the calculation. An SSH service or low-volume application has a very different profile from a popular game server or a media service. Compare included traffic and overage terms with the expected workload. The objective is not to find the lowest theoretical address price. It is to find the smallest stable monthly setup that can carry the service.

Keep the server, move only the public route

A routed IPv4 separates the address used by visitors from the physical location of the server. Visitors connect to the public /32. Taipan receives that traffic on its network and carries it through the selected tunnel to the customer server. The server can remain at home, in an office, in a rack, or with another host. The customer does not need an ASN, a BGP session, or ownership of an IPv4 block for this one-address design.

WireGuard is a practical delivery method when the existing server is not at a fixed network location. The server establishes the tunnel outward, and the public address remains the same address that visitors use. If the local connection changes, the tunnel reconnects while the public IPv4 stays at Taipan. This keeps the public side of a small deployment stable without requiring the workload to move.

GRE and IPsec are available when the endpoint and existing equipment suit those protocols better. The choice should follow the network and the server, not a generic feature list. What matters to a buyer of one cheap IPv4 is that the address reaches the right machine and that the path remains understandable to operate.

Start small and avoid the wrong purchase

Buying IPv4 space outright is not an economical answer to a one-address need. Ownership transfers are generally associated with larger blocks and add registry, routing, and capital costs. Renting one routed address keeps the commitment aligned with the deployment: one public endpoint, one delivery path, and no unused address inventory.

Start with one address, test the real application from an independent network, and add capacity only when the service creates a specific need. This avoids paying for addresses because they seem cheap while preserving a clean path to expand later. The guide to routed /29, /28, and /27 IPv4 subnets explains when a larger allocation becomes a justified operational choice.