Static vs dynamic IP addresses

Check the address you currently hold on the full report, then compare it after a reboot.

Two different questions with the same name

“Static IP” means one thing inside your network and another thing on the internet side. Keeping them apart avoids most of the confusion.

	Inside your LAN	On the internet side
Address	192.168.1.50 and similar	The public address on this site
Who assigns it	Your router’s DHCP server	Your ISP
Can you set it	Yes, in the device or as a DHCP reservation	No, only the ISP can
Why you would	So a printer or NAS keeps the same address	So a server keeps the same public address
Cost	Nothing	Usually a monthly fee, often business-tier only

A “static IP” article that tells you to open network settings and type an address is talking about the LAN side. That does not change what websites see.

How a DHCP lease actually behaves

DHCP is defined by RFC 2131. A lease has a duration, and the client tries to renew at half of it. If the renewal succeeds, and it normally does, the client keeps the same address.

That renewal behaviour is why a dynamic address is stable in practice. Typical ISP lease times run from a few hours to a week, and the server prefers to hand back the address a client already had. Your address usually changes only when something interrupts that cycle:

• The modem or router stays powered off long enough for the lease to expire and the address to be reassigned.
• The ISP renumbers a segment, migrates you to different equipment, or performs maintenance.
• You change the MAC address the ISP sees, normally by swapping the router.
• The ISP deliberately rotates addresses, which some do on a fixed schedule.

The result is a spectrum rather than two categories. Many cable and fibre customers hold the same “dynamic” address for a year.

Where each type is used

Dynamic is the default for residential broadband and for every mobile connection. It costs the ISP nothing extra and lets them recycle addresses between customers, which matters now that IPv4 space is bought rather than allocated.

Static is standard on business lines and on server hosting. Providers sell it as a single address or a small block, often a /29 that gives you five usable addresses.

Neither applies if you are behind carrier-grade NAT. Your router holds an address from 100.64.0.0/10 and the public address is shared with other subscribers. It may change several times a day and you cannot receive inbound connections on it at all.

IPv6 changes the shape again. Your ISP delegates a prefix rather than one address, and the prefix may be static even when the IPv4 address is not, or the reverse. Check both on /ipv6-test.

When you genuinely need a static public IP

• You run a mail server. Receiving mail needs a fixed address with matching forward and reverse DNS, and most large providers reject mail from residential dynamic ranges regardless.
• Someone else’s firewall allows your address. Corporate VPNs, database allowlists, and payment gateways often permit a specific IP. A change locks you out.
• You hold a licence tied to an address. Some enterprise software and camera systems bind to one.
• You need consistent reverse DNS. Only the address holder can set a PTR record, and ISPs set them on static assignments.

What does not need a static address: hosting a game server, reaching your home NAS, running a camera, or exposing a small website. Dynamic DNS covers all of those.

Dynamic DNS as the cheaper answer

A dynamic DNS service gives you a hostname and a small client that reports your current address whenever it changes. The DNS record follows you, so home.example.net keeps resolving.

Most consumer routers include a DDNS client. ASUS, Synology, and Ubiquiti ship their own; almost every router supports at least one third-party provider. If yours does not, run the updater on a machine that is always on.

Two practical notes. Set the record’s TTL low, 60 to 300 seconds, so resolvers pick up a change quickly; see how TTL works. And DDNS does nothing behind CGNAT, because the address it publishes is not reachable in the first place.

How to ask your ISP

Static addresses are sold, not configured, so this is a billing conversation:

Is a static public IPv4 address available on my current plan, and what is the monthly cost? If not, which plan includes one?

Expect one of three answers: a small monthly fee added to your existing line, a required upgrade to a business plan, or a flat no on consumer service. Mobile carriers almost never offer one.

Before paying, confirm you are not behind CGNAT, because a static address on a CGNAT line is not something an ISP can sell you without also moving you off the shared translator. The check takes a minute: compare your router’s WAN address with the address on this site.