IPv6 capability now past 50 per cent in APAC region as well
European region still well behind Asia-Pacific and northern America
European region still well behind Asia-Pacific and northern America
More than half of internet users in the Asia-Pacific region are now ‘IPv6 capable’ according to APNIC, the region’s RIR. India leads the way, with capability at 78 per cent. That very high figure is a consequence of India’s historical shortage of IPv4 addresses, central government policy and a very large end user population.
Asia-Pacific is the second region where IPv6 capability is now more than 50 per cent. Northern America (the region overseen by ARIN) announced it was past that threshold a while ago, and now boasts 52 per cent capability. As the following table shows, the regions overseen by LACNIC (southern America) and RIPE NCC (greater Europe) have much lower levels of IPv6 capability, while capability in Africa (AFRINIC) is negligible.
RIR | regio | IPv6-capability |
---|---|---|
ARIN | Northern America | 52% |
APNIC | Azia-Pacific | 50% |
LACNIC | Southern America | 39% |
RIPE NCC | Greater Europe | 28% |
AFRINIC | Africa | 4% |
Figure 1: : IPv6 capability in the APAC region passes 50 per cent. [Source: APNIC]
Because it includes India and China, the APNIC’s Asia-Pacific (APAC) region has far more end users than any other region. Not surprisingly, therefore, Google’s data indicates that global IPv6 adoption is broadly similar to the APAC level. Nearly half (48 per cent) of all users currently reach Google via an IPv6 connection. That figure has been steadily edging up for some years at a little less than 5 percentage points a year.
Figure 2: Global client-side IPv6 adoption as measured by Google, July 2025. [Source: Google]
According to Google, adoption in the Netherlands is now 36 per cent – well below the levels in neighbouring countries. Our disappointing level of adoption is more on a par with countries on the periphery of Europe.
Even though we’ve known about the problem for years, and despite the fact that it holds back our economy, IPv6 adoption in the Netherlands is only falling further behind.
Figure 3: Dutch client-side IPv6 adoption as measured by Google, July 2025. [Source: Google]
Land | Adoptiepercentage |
---|---|
France | 85% |
Germany | 74% |
Belgium | 68% |
UK | 51% |
Luxembourg | 44% |
Ireland | 39% |
Netherlands | 36% |
Although we’re disappointed about the situation in the Netherlands, we’re much more positive about the adoption of IPv6 more generally. As mentioned above, the number of Google users accessing the company’s services via IPv6 connections continues to grow by roughly 5 percentage points a year.
What's more, it appears that we have passed a significant tipping point. Since 2022, we have witnessed a structural downturn in the price of traded IPv4 address blocks, after years of steady increase. That leads us to suspect that Peak IPv4 is now behind us. As the adoption of IPv6 continues and the importance of IPv4 declines, there must come a point where the value of IPv4 addresses goes into irreversible decline.
Having been roughly 35 dollars per address a year ago, the going rate has dropped towards the 30-dollar mark. For the biggest address blocks, the price drop has been quite dramatic: from nearly 50 dollars per address for a /16 block a year ago to little more than 20 dollars now. So those blocks have lost more than half their value in the space of a year.
We have also entered a new phase in a technical sense. Traditionally, IPv4 has been taken as the basis, and IPv6 has been superimposed or enabled alongside IPv4. Now, when computer networks are built or upgraded, IPv6 should serve as the basis, with support for IPv4-only systems realised by means of an overlaid service.
In that altered landscape, a new group of IPv4-to-IPv6 transition mechanisms has acquired a central role. As we move towards an IPv6-mainly world, it's time for old mechanisms such as dual stack with (CG)NAT to make way for newer alternatives based on NAT64, DNS64 and 464XLAT. For a detailed description of that strategy and the associated technologies, see this article.