IPv4 and IPv6 are both addressing systems for identifying a device on a network, including a web server — the difference that actually matters for hosting isn't speed or security, it's address space, and almost everything practical about this topic follows directly from that one fact.
Why IPv6 exists: address space exhaustion
IPv4 addresses are a 32-bit number, which allows for roughly 4.3 billion unique addresses — a number that seemed effectively unlimited when IPv4 was designed, decades before the internet's current scale. That space has since been fully allocated by the regional registries that distribute it, which is the specific, concrete problem IPv6 exists to solve: its 128-bit address format provides an address space large enough that practical exhaustion isn't a realistic concern in any foreseeable timeframe. IPv6 isn't an upgrade to IPv4 in the sense of replacing something broken — it's a parallel system built because the original one ran out of room. Address scarcity also shows up well before technical exhaustion as cost: the market for leasing or purchasing existing IPv4 address blocks has grown as available supply has tightened, a quiet, ongoing cost pressure feeding into why registries and hosting providers increasingly treat dedicated IPv4 addresses as something to allocate deliberately rather than hand out freely by default.
What actually changed, technically
Beyond the address space itself, IPv6 addresses look structurally different (written in hexadecimal groups separated by colons, rather than IPv4's familiar four decimal numbers separated by dots) and the protocol includes some built-in changes to how address assignment and certain network functions work. For a typical hosted website, none of this changes how the site itself is built or what it serves — it changes how a visitor's device finds and connects to the server, which is a layer below the application itself.
A related symptom: carrier-grade NAT
IPv4 exhaustion shows up in another place worth knowing about, even though it doesn't directly change anything about hosting a site: many internet service providers, facing the same address shortage, now share a single public IPv4 address across many customers using carrier-grade NAT (CGNAT), rather than giving each household its own. This is why a growing number of visitors reach a website through a shared, translated IPv4 address, while their actual device may separately have its own distinct IPv6 address end to end. It's a consumer-network-side symptom of the same underlying scarcity IPv6 was built to solve — not something a hosting setup needs to specifically accommodate, but useful context for why IPv6 adoption keeps gradually increasing from the client side even as IPv4 remains necessary from the server side.
Dual-stack hosting: the common real-world setup
Most hosting today runs dual-stack: a server has both an IPv4 and an IPv6 address simultaneously, and it accepts connections over either one depending on which protocol the connecting visitor's device and network support. This is the practical middle ground between the two protocols — it doesn't require choosing one over the other, and it means a site is reachable by visitors regardless of which protocol their specific connection happens to use. An IPv6-only server, by contrast, is only reachable by visitors whose own network path supports IPv6 end to end, which is a meaningfully more restrictive setup given where IPv4 and IPv6 support actually stand across different networks and regions today.
What it means for a server to have a dedicated IPv4 address
Because the IPv4 address space is fully allocated, dedicated IPv4 addresses are a more constrained resource than they once were, and some hosting plans distinguish between a server having its own dedicated IPv4 address versus sharing one. A dedicated IPv4 address matters specifically for things that depend on a server's address being uniquely its own — certain SSL certificate configurations, some email-sending reputation considerations, and any situation where IP-based access rules need to identify one specific server unambiguously. IPv6 address space isn't under the same constraint, which is part of why dual-stack hosting (rather than IPv6 exclusively) remains the practical default — it preserves dedicated-IPv4 use cases while still providing the much larger IPv6 space. Some hosting plans now offer IPv4 as a paid add-on rather than a default inclusion, specifically reflecting that underlying scarcity, while IPv6 allocation remains comparatively generous and typically included without extra cost.
The DNS records involved: A and AAAA
At the DNS level, an IPv4 address is published through an A record, and an IPv6 address through an AAAA record — both can exist for the same hostname simultaneously on a dual-stack setup, and a visitor's own device and network determine which one actually gets used for a given connection. See DNS Records, Nameservers and Resolution for how these record types fit into DNS more generally — this article assumes that background rather than re-covering every record type here.
Does a given site actually need IPv6?
For most business websites, the practical answer is that IPv6 support is a reasonable thing to have (and increasingly common as a hosting default) rather than an urgent requirement to actively pursue — dual-stack hosting provides it without extra effort on the site owner's part in most cases, since it's typically a hosting-environment-level configuration rather than something built into the website's own code. A site actively excluding IPv6-only visitors would be a real problem, but that scenario is uncommon precisely because dual-stack hosting has become the standard default rather than something that has to be deliberately requested.
Common misconceptions
A few assumptions about this topic are worth correcting directly. IPv6 is not inherently faster than IPv4 — any performance difference in a specific case comes from routing and network path differences, not from the protocol itself having a speed advantage. IPv6 is also not inherently more secure — it has some built-in features that older IPv4 deployments sometimes lacked by default, but a server's actual security posture depends on its own configuration regardless of which protocol is in use, not on IPv6 providing automatic protection. And having an IPv6 address does not mean a server is somehow more "modern" in a way that affects anything but reachability — it's an addressing detail, not a measure of a hosting setup's overall quality.
The practical reality: IPv4 reachability still generally matters
Overall IPv6 adoption has grown steadily over time, but a meaningful share of networks, devices and intermediate infrastructure still rely on IPv4 today, and that share varies significantly by region and network type rather than following one global number — which is exactly why this article won't cite a specific current adoption percentage as a fixed fact. The practical consequence for hosting is straightforward regardless of the exact current figure: a site generally still needs IPv4 reachability to serve its full potential audience, and dual-stack hosting is what provides that without requiring a site to choose one protocol and exclude visitors using the other.
Why this usually isn't a decision you have to make
For almost every business website, dual-stack hosting — provided by default on most modern hosting, including VPS hosting with root-level network configuration access — already handles this correctly without requiring an active choice between IPv4 and IPv6. The two protocols aren't competing options for a typical site; they're both already running together, and the underlying reason IPv6 exists at all (IPv4's address space running out) is a registry-and-infrastructure-level problem dual-stack hosting has already absorbed on the site owner's behalf.
