Views: 1,151

The lazy IPv4 take in 2026 is that softer pricing means demand cooled off. That is the kind of conclusion you reach when you watch price alone and ignore throughput, buyer mix, and registry workflow. The better read is more interesting. Demand is still there. It is just showing up in a market that is broader, less theatrical, and a lot more disciplined than the hyperscaler-fueled frenzy people got used to talking about. That is why this month’s story should not start with panic or nostalgia. It should start with mixed signals: strong enough annual throughput to keep the market relevant, monthly demand that looks steady rather than euphoric, and VPS-heavy buyer activity that says IPv4 is still an operating asset, not a museum piece. That matters because the market does not need to look euphoric to be healthy. Market context still favors a story of durable throughput, softer pricing, and a buyer base that is wider than the usual hyperscaler talking points suggest.

Hostinger and Hetzner Matter for a Reason

The top-recipient table is useful this month because it points away from the tired idea that only giant cloud buyers still matter. Hostinger and Hetzner sit inside the hosting and VPS infrastructure world, and that matters because their buying behavior reflects operational demand, not vanity accumulation. These companies consume IPv4 as part of actively routed service footprints tied to hosting, cloud instances, customer onboarding, and BGP-announced production infrastructure.

Hostinger with 720k IPs and Hetzner 400k IPs purchased this year provide the VPS-provider trend in plain view. That segment is still buying, and the likely reasons are not mysterious: more cloud-cost sensitivity, more demand for flexible infrastructure, and more customers who still need IPv4-first compatibility. AI demand adds pressure around that ecosystem even when the biggest AI buyers are not the exact names sitting in the VPS table.

None of that means the top of the market stopped being concentrated. Amazon is still the scale benchmark. With 20 Million purchased IPv4 addresses, we are in route to hit 70 to 80 Million IPs transferred in 2026 setting a new record. That matters because Hostinger and Hetzner tell you the market is broadening, but Amazon reminds you it is not suddenly flat or democratic. A few very large buyers still shape scarcity psychology for everyone else.

That is the real mixed-signals layer. The market is broadening beyond hyperscalers, but it is still not free from top-heavy influence. Smaller infrastructure buyers may be more active, but they are not moving inside a vacuum.

ARIN Demand Signals Bull Market

The ARIN transfer-request count for the prior month came in at 174 requests, and that matters because it is a direct demand reading rather than a pricing proxy. Based on the 2026 request pattern in the research packet, that level sits above both the current-year average monthly pace and the 3-year average monthly of 150 requests per month. In other words, the request pipeline is not just active. It is running ahead of both the shorter and longer baselines that matter for market timing.

That becomes more meaningful when it is read next to annual transfer pace and large-buyer behavior. If requests are still running hot while annual transfer volume is already pacing toward a standout year, then the better interpretation is not that supply is getting easier. It is that buyer intent is broadening while cleaner inventory is still being absorbed. That is not a blowout month, but it’s a signal that demand is ahead of the bear market norms.

AI and BEAD Keep the Floor in Place

AI matters because it pushes more demand into cloud, hosting, edge, and interconnect environments that still rely on IPv4 in production. It does not need to recreate 2021 to 2022-style frenzy to matter. It only needs to keep infrastructure buyers active, and it is doing that. The result is a market where flexible, fast-turn consumption can stay strong even when buyers are more careful about price.

BEAD matters for a different reason. It creates deployment pressure. Once more operators move from planning to actual builds, IPv4 is still needed for subscriber onboarding, management systems, CG-NAT, and mixed-stack service delivery.

What Comes Next

For the next quarter, expect stable to slightly firmer demand rather than a dramatic breakout. This year still looks like a market with healthy transaction volume and more disciplined pricing. Into next year, the most likely outcome is not a panic spike. It is a sturdier pricing floor, persistent small-block demand, and continued interest from hosting, cloud-adjacent, and broadband-driven buyers.

That is why the lazy “prices softened so demand disappeared” story fails. The better story is that IPv4 demand is broader than it looks, the buyer mix is getting more interesting, and the market is behaving like a real infrastructure market instead of a hype cycle.

Other Popular Blog Posts

The Importance of Prefix Diversity in IPv4 Addressing

ARIN Waiting List For IPv4 Still Over a Year in 2026

Google & Meta Take Control of Undersea Backbone