A static public IP address is a single internet address that belongs to you, doesn't change, and is reachable from anywhere on the internet. It is what makes it possible to host anything at a location other people need to connect to — a web server, an email server, a VPN, a Minecraft server, a Home Assistant instance, a security-camera feed. Without one, the rest of the internet has no stable way to reach whatever you're running.
Most home and small-office connections don't come with one. This page explains why that's changed over the last decade, what your options actually are in 2026, and the real-world trade-offs of each one. If you're looking for a step-by-step walkthrough on the easiest way to get one, the how-to-get-a-static-ip-address guide covers it in detail.
Why a truly static, truly public IP is rarer than it used to be
In 2010 you could plug a server into most home broadband connections and the world could reach it. That stopped being true for a specific reason: there are no IPv4 addresses left. The IPv4 address space tops out at about 4.3 billion unique addresses, and they have all been allocated to regional internet registries since 2011. New addresses now trade on a secondary market at roughly $40–$60 per IP, which means giving a single customer a dedicated address costs your ISP a real capital expense they'd rather avoid.
Instead, ISPs do two things:
- Dynamic allocation: they hand out IPs from a small pool, recycling them as customers disconnect. Your address might stay the same for weeks, then change without warning.
- CGNAT: they share a single public IP across hundreds or thousands of customers by NATting at the carrier edge. You never see a real public IP at all — you get a 100.64.x.x address instead — and inbound connections are effectively impossible.
Both of these are fine for browsing, streaming, or video calls, which only need outbound connections to work. Neither is usable for self-hosting anything that the outside world needs to reach. The result: if you want a static public IP in 2026, you have to go look for one.
The four real options
| Option | Monthly cost | Portable across locations | Runs on your own hardware | Typical email deliverability |
|---|---|---|---|---|
| Request one from your ISP | $10–$100+ (usually business plan) | No — tied to the physical line | Yes | Poor (residential CIDR often blocklisted) |
| Rent a VPS (DigitalOcean, Hetzner, etc.) | $5–$50 | Yes, but requires moving workloads to the VPS | No — workload runs in the cloud | Good (after warmup) |
| Dedicated public-IP service (e.g. GetPublicIP) | $8.99/mo per IP | Yes — IP travels with the account | Yes — tunnel terminates on your hardware | Good (commercial IP ranges) |
| Mesh VPN with static exit (Mullvad, Tailscale Funnel) | $5–10 | Partly — exit node IP stays yours while subscribed | Yes | Varies — often poor (shared VPN reputation) |
None of these is strictly better than the others — the right one depends on what you're optimising for.
ISP static IP
Works well if you have a business-tier connection already, don't care about email, and never plan to move. Often the cheapest option if you're already paying for a business line. Dealbreakers: tied to one address, usually still residential from the internet's point of view (blocklists), and ISP customer support for "I'd like to open port 25" calls ranges from slow to hostile.
VPS
Makes sense if your workload can actually live in the cloud: a blog, a bot, a static site. Stops making sense if the point is to run something on your own hardware at home — there, you're paying for a VPS and your home server and still don't have inbound connectivity at home. A VPS is also the de facto answer for anyone who needs a clean commercial IPv4 for email.
Dedicated public-IP service
A service like GetPublicIP gives you a real public IPv4 (and IPv6) on dedicated infrastructure, then tunnels traffic to your home or office server over WireGuard. Inbound packets arrive on the public IP, travel through the encrypted tunnel, and land on your server. Outbound replies flow the same way. The IP is portable — move house, change ISPs, switch to 5G as a failover — and the service provider is the one dealing with upstream blocklists, abuse desks, and IPv4 scarcity. Think of it as "the useful half of a VPS" without moving your workload to someone else's computer.
Mesh VPN with static exit
Tailscale Funnel, Cloudflare Tunnel, and Mullvad's static IP plans all technically let people on the internet reach your machine. Each has its own limits: Cloudflare Tunnel terminates SSL on Cloudflare's edge (they see your traffic in plaintext), Tailscale Funnel only works for HTTPS and only for allowlisted domains, and Mullvad-style VPN static IPs are often shared with other customers. Fine for quick one-offs; weaker than the first three options for anything long-running.
Cost and trade-off summary
- Cheapest overall: ISP static IP on an existing business plan, if you can get one and don't care about email or portability.
- Cheapest for portability: Dedicated public-IP service — the $8.99/mo pays for the IP and for the fact that you're not buying it from your ISP.
- Best for clean email: VPS or dedicated public-IP service — anything routed through a known commercial range.
- Worst value: Mesh VPN static-IP add-ons — low price but heavy limitations that catch up to you.
The main question to ask before picking one is "how stable is this physical line?" If the answer is "I might move, change ISPs, or want to failover to 5G during an outage," then anything tied to your physical connection is a trap — it's cheaper right now but forces a migration later. Services where the IP lives with your account, not your line, avoid that.
When a static IP isn't actually what you need
Not every self-hosting problem needs a static public IP. If your only use case is remote access to your own services — i.e., you're the only user — a mesh VPN like Tailscale or WireGuard is simpler, free for small usage, and doesn't require a public IP at all. Static IPs earn their cost when other people need to reach your services: customers visiting a website, mail servers delivering email, game clients connecting to a server, remote IoT devices phoning home.
Most of the time, the question "do I need a static IP?" really means "do I need other people on the internet to be able to reach something I'm running?" If yes, you need one. If no, there's probably a VPN answer that's cheaper and easier.
Related guides
- How to get a static IP address — step-by-step walkthrough
- How to get a public IP address — focused on the "public" part of the question
- Self-hosting behind CGNAT — if your ISP is the blocker
- Getting started with GetPublicIP — if you've decided on the dedicated-service route