IPv4 uses 32-bit addresses, about 4.3 billion in total, and the free pools of ARIN, the RIPE NCC and LACNIC are empty. New IPv4 comes from the transfer market, at about $14 to $24 per address in August 2026 depending on block size, or from a lease at about $0.35 per address per month.
IPv6 uses 128-bit addresses and is not scarce. About 47% of users reached Google over IPv6 in mid-September 2026. Most networks run both protocols, so the planning question is how fast the IPv4 side can shrink without disrupting production.
Understanding the Current State
IPv4: Scarcity Becomes Reality
IPv4 ran out years ago. IANA handed out its last five /8s on 3 February 2011, and the registries then reached their final phase one by one: APNIC (April 2011), the RIPE NCC (September 2012), LACNIC (June 2014) and ARIN (24 September 2015). The RIPE NCC ran out completely on 25 November 2019. On 25 September 2026, 791 LIRs were on its waiting list, the first of them waiting 485 days, and each LIR can only ever receive one /24 (ripe-826). ARIN had 514 approved requests on its waiting list on the same day, the oldest from June 2025, at most a /22 each. APNIC gives at most a /23 per member. AFRINIC still had about 2,384 /24s available on 24 September 2026, handed out /24 to /22 at a time. The full story is in IPv4 exhaustion: why leasing matters.
IPv4 market in August 2026:
- Purchase price: about $23.50 per address for /22 to /24 blocks, about $17 for /20 to /21, about $14 for /17 and larger (monthly averages reported by broker IPv4.Global, online and private sales mixed, priced at agreement date). A /24 therefore costs about $6,000.
- Supply: transfers only. In the RIPE region alone 5,935,616 IPv4 addresses changed hands in July 2026 (RIPE NCC member update, August 2026).
- Leasing: about $0.35 per address per month on average on the IPXO platform (last six months, data updated 25 September 2026).
- Block size: small blocks carry the premium. In August 2026 a /22 to /24 cost about $23.50 per address against about $14.30 for a /16 or larger, and /16+ blocks had caught up with /17 to /19 after bottoming in January 2026.
IPv4 addresses now trade like a commodity, so an IPv4 plan is also a budget line.
IPv6: Crossing the Majority Threshold
On 28 March 2026 the share of users reaching Google over IPv6 passed 50% for the first time (50.10%), and it peaked at 50.91% on 18 July 2026 (Google IPv6 statistics). Weekends run higher than weekdays: in September 2026 Saturdays and Sundays measured 48.7% to 50.6%, ordinary weekdays about 46%. The weekly average was 47.4% for 14 to 20 September 2026, against 46.6% a year earlier and 43.6% in 2024.
IPv6 adoption by country (Google, native IPv6, 20 September 2026):
| Country | IPv6 share |
|---|---|
| France | 86.3% |
| Germany | 75.4% |
| India | 68.5% |
| Japan | 56.9% |
| Brazil | 56.3% |
| United States | 55.5% |
| United Kingdom | 52.4% |
| China | 8.1% |
| Global, Google (7-day average, 14 to 20 September 2026) | 47.4% |
| Global, APNIC Labs (23 April 2026) | 42% |
Sources: Google per-country statistics and APNIC Labs. The two differ because Google counts its own users' connections while APNIC measures capability through ad-based tests, and Google has almost no users in China (APNIC blog, April 2026). Country figures move week to week. In the RIPE NCC region, 15,806 of the 20,687 LIR accounts held an IPv6 allocation in August 2026.
Content networks, mobile operators and the large clouds all serve IPv6, so an IPv4-only network reaches them through NAT.
The Dual-Protocol Reality
For the foreseeable future, networks must support both IPv4 and IPv6. Pure IPv6-only deployment remains impractical for most organizations due to:
- Legacy application dependencies
- Third-party service integrations
- Customer connectivity requirements
- Equipment limitations
- Operational expertise gaps
Plan for years of dual-stack, not months.
IPv4 vs IPv6: What Is the Difference?
Address Space
IPv4:
- 32-bit addresses (4.3 billion total)
- Notation: 192.0.2.1
- Exhausted globally
- Requires NAT for address conservation
IPv6:
- 128-bit addresses (340 undecillion addresses)
- Notation: 2001:db8::1
- Virtually unlimited
- Eliminates need for NAT
Network Configuration
IPv4:
- DHCP for automatic configuration
- Manual configuration common
- Address conservation techniques required
- Complex NAT configurations
IPv6:
- SLAAC (Stateless Address Autoconfiguration)
- DHCPv6 for managed configuration
- Simplified routing with aggregation
- End-to-end connectivity restored
Security Considerations
IPv4:
- IPsec optional
- Mature security tools and practices
- Well-understood threats
- NAT provides obscurity (not security)
IPv6:
- IPsec support recommended (SHOULD) for every IPv6 node by RFC 8504; it was a MUST until RFC 6434 in 2011, so in practice it is optional as in IPv4
- Larger attack surface with abundant addresses
- Evolving security practices
- New monitoring challenges
Performance Characteristics
Performance depends on the path more than on the protocol. RFC 8305 (Happy Eyeballs version 2) makes dual-stack clients try IPv6 first and fall back to IPv4 within a fraction of a second, so a broken IPv6 path costs a short delay rather than a failed connection. Where IPv6 removes carrier-grade NAT from the path, it removes a stateful hop.
Strategic Planning Framework
Step 1: Assess Current State
IPv4 Inventory Audit
Document your complete IPv4 footprint:
- Registered holdings: allocations and assignments registered to your organization at an RIR (the ASN lookup lists the prefixes your AS announces today)
- Provider-assigned addresses: PA space from ISPs
- Leased addresses: Temporary holdings
- Utilization rates: Actual usage vs. allocated
- Critical dependencies: Systems requiring public IPv4
Calculate your IPv4 burn rate and projected needs for 3-5 years.
IPv6 Readiness Assessment
Evaluate your organization's IPv6 maturity:
Infrastructure:
- Network equipment IPv6 capability
- Server/application IPv6 support
- Security infrastructure (firewalls, IDS/IPS)
- Monitoring and management tools
- Load balancers and proxies
Applications:
- Business-critical applications IPv6 compatibility
- Third-party service dependencies
- Custom application code review needed
- Database systems IPv6 readiness
Personnel:
- IPv6 knowledge and training levels
- Operational experience with IPv6
- Security team IPv6 expertise
- Development team capabilities
Step 2: Define Requirements
Short-Term Needs (0-18 Months)
- Immediate IPv4 requirements for production
- Customer-facing service dependencies
- Compliance and regulatory obligations
- Critical system upgrades or expansions
Medium-Term Needs (18-36 Months)
- Planned infrastructure expansion
- Application modernization roadmap
- Market expansion or geographic growth
- Technology refresh cycles
Long-Term Vision (3-5 Years)
- Strategic architecture goals
- IPv6-native infrastructure targets
- IPv4 dependence reduction timeline
- Cost optimization objectives
Step 3: Evaluate Options
The four IPv4 options come first, then the three IPv6 deployment strategies.
IPv4 Strategy Options
Option 1: Purchase IPv4 Addresses
When This Makes Sense:
- Long-term IPv4 requirements certain
- Capital budget available (about $23.50 per address for /22 to /24 blocks, August 2026)
- Accounting treatment as capital asset acceptable
- Organizational stability and asset management capability
Advantages:
- Registered to your organization for as long as you meet the RIR's policies and pay its fees
- No recurring lease cost
- Resale value (prices move both ways: about $27.40 in September 2025, $23.50 in August 2026 for /22 to /24)
- Full control over routing and RPKI
Disadvantages:
- High upfront cost (about $6,000 for a /24 and about $940,000 for a /16 at August 2026 averages)
- Ongoing RIR fees
- Asset management responsibilities
- Market value exposure
Financial example:
For a /22 (1,024 addresses) at about $23.50 per address (IPv4.Global, August 2026), the block costs about $24,000. At ARIN, add the $500 transfer fee and the $375 recipient fee for a /22, then $550 a year on the 2X-Small plan (2026 fee schedule, $577.50 in 2027). At the RIPE NCC the transfer itself is free, but a transferred allocation has to sit in an LIR account: €1,800 a year under the RIPE NCC Charging Scheme 2026 (ripe-848) if you are not already a member. A PI block bought on the market avoids that: it can be registered through a sponsoring LIR instead, at €75 a year to the RIPE NCC, €135 a year through our sponsorship (as of September 2026, setup on quote). The block keeps a resale value: for this size the average went from about $27.40 in September 2025 down to about $21.60 in April 2026 and was back near $23.50 in August 2026. RIR fees compared lists the other registries.
Option 2: Lease IPv4 Addresses
When This Makes Sense:
- Temporary or uncertain IPv4 needs
- Limited capital budget
- Operational expense treatment preferred
- Flexibility matters more than holding the registration
Advantages:
- Lower upfront cost
- Flexibility to scale up or down
- OpEx accounting treatment
- No long-term asset liability
Disadvantages:
- Ongoing monthly costs
- Nothing to resell at the end
- Lease availability and pricing volatility
- Contract dependencies
Financial example:
For the same /22 at $0.35 per address per month (IPXO platform average, September 2026): about $358 a month, $4,300 a year and $21,500 over five years. Against the ARIN purchase above (about $24,900 up front, then $550 a year), leasing stays cheaper for about six and a half years, before counting what the bought block could be resold for. The detailed numbers are in IPv4 leasing vs purchasing: cost analysis.
Leasing makes financial sense for organizations expecting to reduce IPv4 dependency within 5 years.
Option 3: IPv4 Conservation Techniques
When This Makes Sense:
- Budget constraints limit acquisition
- Existing allocations underutilized
- Time needed for IPv6 transition
- Internal optimization possible
Techniques:
- NAT optimization: Improve port allocation, connection tracking
- IPv6 migration: Move internal services to IPv6
- Address reclamation: Identify and recover unused addresses
- Segmentation: Optimize subnet allocation
- Load balancer optimization: Share addresses across services
An address audit usually finds space that is assigned but unused; reclaim it before you buy.
Option 4: IPv6-Primary with IPv4 as a Service
When This Makes Sense:
- Modern infrastructure and applications
- Greenfield deployments
- Cloud-native architecture
- Forward-looking strategy
Approach:
- Deploy IPv6 natively for all internal services
- Use NAT64/DNS64 for IPv4 compatibility
- Maintain minimal IPv4 for external-facing services
- Rent public IPv4 from the cloud provider only where needed: AWS charges $0.005 per public IPv4 address per hour since 1 February 2024, about $43.80 per address per year
IPv6 Deployment Strategies
Strategy 1: Dual-Stack Deployment
Description: Run IPv4 and IPv6 simultaneously across all infrastructure, allowing gradual transition without service disruption.
Implementation Approach:
- Phase 1: Infrastructure Preparation
- Upgrade network equipment firmware
- Enable IPv6 on core routing
- Configure IPv6 on switches and routers
- Implement IPv6 addressing scheme
- Phase 2: Service Enablement
- Deploy IPv6 to DMZ services
- Enable dual-stack on web servers
- Configure DNS AAAA records
- Implement monitoring for both protocols
- Phase 3: Internal Migration
- Roll out IPv6 to internal networks
- Migrate internal services
- Train staff on dual-stack operations
- Optimize routing policies
- Phase 4: Optimization
- Prefer IPv6 where possible
- Monitor protocol usage ratios
- Identify IPv4 dependencies
- Plan IPv4 reduction
Advantages:
- Gradual, low-risk transition
- Maintains full backward compatibility
- Flexible rollback options
- Proven approach widely deployed
Disadvantages:
- Two protocols to manage
- Higher operational overhead
- Doubled address management
- Potential for configuration drift
Strategy 2: IPv6-Only with Translation
Description: Deploy IPv6-only infrastructure with translation mechanisms (NAT64/DNS64) to access IPv4 resources.
When This Makes Sense:
- New deployments without IPv4 legacy
- Cloud-native applications
- Mobile networks
- Data centers with modern workloads
Implementation Requirements:
- NAT64 gateways for IPv4 connectivity
- DNS64 for automatic translation discovery
- 464XLAT for applications requiring IPv4 literals
- Monitoring for translation performance
- RFC 8925 (IPv6-Only Preferred option for DHCPv4) where hosts should decline IPv4 on a dual-stack segment, the basis of IPv6-mostly networks
Advantages:
- Eliminates IPv4 management overhead
- Simplifies addressing schemes
- Lower long-term costs
Disadvantages:
- Translation overhead and complexity
- Some applications incompatible
- Troubleshooting challenges
- Less operational experience in industry
Strategy 3: Tunneling Solutions
Description: Tunnel IPv6 traffic over IPv4 infrastructure (or vice versa) as transitional mechanism.
Common Tunneling Approaches:
- 6to4 (legacy): Automatic IPv6 over IPv4 tunneling
- 6PE/6VPE: IPv6 across an MPLS core (RFC 4798, RFC 4659)
- Teredo (legacy): IPv6 connectivity through NAT
- ISATAP (legacy): Intra-site IPv6 transport
6to4 anycast relays were deprecated by RFC 7526 in 2015, which found the mechanism unsuitable for widespread use and told vendors to stop supporting it. Teredo and ISATAP date from the same era. For new work, 6PE/6VPE carries IPv6 across an MPLS core and translation (NAT64, 464XLAT) replaces host tunnels.
When This Makes Sense:
- Isolated IPv6 needs in IPv4 infrastructure
- Connecting IPv6 islands
- Temporary solution during transition
- Specific application requirements
Disadvantages:
- Adds complexity and overhead
- Performance penalties
- Security considerations
- Should be temporary solution only
Recommended Transition Roadmap
Phase 1: Foundation (Months 0-6)
Objectives:
- Complete IPv4/IPv6 assessment
- Develop a transition plan
- Secure budget and resources
- Begin team training
Key Activities:
- IPv4 inventory and utilization audit
- IPv6 readiness assessment
- Equipment compatibility verification
- Address allocation planning
- Stakeholder alignment
Deliverables:
- IPv4/IPv6 strategic plan document
- 3-year roadmap with milestones
- Budget proposal
- Risk assessment
Phase 2: Infrastructure Preparation (Months 6-12)
Objectives:
- Enable IPv6 on network infrastructure
- Establish dual-stack capability
- Implement monitoring and security
Key Activities:
- Network equipment IPv6 enablement
- IPv6 space: a PA block from your provider or LIR, or a PI /48 registered to your organization (PI vs PA resources); LIRs receive a /32 to /29 allocation under ripe-738
- DNS infrastructure IPv6 configuration
- Firewall and security policy updates
- Monitoring tool IPv6 support
Deliverables:
- IPv6-capable network infrastructure
- Dual-stack addressing scheme implemented
- Security policies updated for IPv6
- Monitoring dashboards for both protocols
Phase 3: Service Migration (Months 12-24)
Objectives:
- Migrate customer-facing services to dual-stack
- Deploy IPv6 to priority applications
- Establish operational procedures
Key Activities:
- Web services dual-stack deployment
- DNS AAAA record configuration
- Load balancer IPv6 configuration
- Application-by-application migration
- Performance monitoring and optimization
Deliverables:
- Major services IPv6-accessible
- Operational runbooks for dual-stack
- Performance baseline established
- User experience validation
Phase 4: Optimization and IPv4 Reduction (Months 24-36)
Objectives:
- Increase IPv6 traffic ratio
- Reduce IPv4 dependency
- Optimize costs
Key Activities:
- Prefer IPv6 in routing policies
- Migrate internal services to IPv6
- IPv4 address reclamation
- NAT64/DNS64 deployment for new services
- IPv4 allocation reduction planning
Deliverables:
- 50%+ traffic over IPv6
- Reduced IPv4 allocation needs
- Cost optimization achieved
- Clear path to IPv6-primary architecture
Phase 5: IPv6-Primary (Months 36+)
Objectives:
- Operate primarily on IPv6
- Maintain minimal IPv4 for compatibility
- Achieve cost and complexity reduction
Key Activities:
- IPv6-only internal networks
- IPv4-as-a-Service for external requirements
- Legacy application remediation
- Staff expertise development
Deliverables:
- IPv6-native architecture
- Minimal IPv4 footprint
- Operational efficiency gains
Decision Framework: Key Questions
Financial Considerations
- What is our 5-year budget for IP addresses?
- Capital vs. operational expense preference
- Total cost of ownership calculations
- IPv4 market price exposure
- What is our IPv4 utilization efficiency?
- Current usage vs. allocation
- Reclamation opportunities
- Conservation potential
- What are our IPv6 deployment costs?
- Equipment upgrades required
- Training investments
- Operational overhead
- External consulting needs
Technical Considerations
- What is our application IPv6 readiness?
- Critical applications compatible?
- Third-party dependencies identified?
- Custom code remediation needed?
- What is our infrastructure IPv6 capability?
- Network equipment support
- Security infrastructure readiness
- Monitoring tool compatibility
- What are our performance requirements?
- Latency sensitivity
- Geographic distribution needs
- Scale and growth projections
Business Considerations
- What is our risk tolerance?
- Service disruption acceptance
- Complexity management capability
- Vendor lock-in concerns
- What are our timeline constraints?
- Business growth plans
- Technology refresh cycles
- Compliance deadlines
- What is our competitive positioning?
- Industry IPv6 adoption rates
- Customer expectations
- Market differentiation opportunities
Industry-Specific Considerations
Internet Service Providers (ISPs)
Key Priorities:
- IPv6 deployment to customers (CGN avoidance)
- Large-scale address management
- Regulatory compliance
- Customer equipment support
Recommended Approach:
- Dual-stack deployment to subscribers (see IPv6 leasing benefits for ISPs)
- IPv6-only core network (optional)
- DS-Lite or MAP-E for IPv4 continuation
- Gradual customer migration
Enterprise Networks
Key Priorities:
- Application compatibility
- Security policy enforcement
- Gradual migration without disruption
- Cost control
Recommended Approach:
- Dual-stack internal networks
- Phased application migration
- IPv6-capable security infrastructure
- Long-term coexistence planning
Cloud-Native Organizations
Key Priorities:
- Modern architecture
- Minimal IPv4 dependency
- Cost optimization
- Scalability
Recommended Approach:
- IPv6-primary architecture
- NAT64/DNS64 for IPv4 access
- Cloud provider IPv6 services
- Minimal IPv4 allocation
Mobile Network Operators
Key Priorities:
- Address efficiency for massive device counts
- Performance optimization
- Standards compliance
Recommended Approach:
- IPv6-only mobile networks
- 464XLAT for application compatibility
- Carrier-grade NAT for legacy IPv4
- Strategic IPv4 retention for services
Common Pitfalls to Avoid
Pitfall 1: Delaying IPv6 Indefinitely
Every year without IPv6 adds equipment, applications and staff habits that will have to be migrated later, at a higher price. Start the planning now even if the deployment runs over several years.
Pitfall 2: Underestimating Complexity
IPv6 touches every layer: routing, security policies, monitoring, applications and training, and each of them is usually underestimated. Assess before you commit to a timeline and add a contingency line to the budget.
Pitfall 3: Treating IPv6 as "IPv4 with More Addresses"
IPv6 operates differently, and a design that copies IPv4 habits (small subnets, NAT at the edge) misses the simpler architecture it allows. Budget for IPv6-specific training and an architectural review.
Pitfall 4: Ignoring Security Implications
Firewalls, IDS/IPS and monitoring must handle IPv6 explicitly, and the larger address space changes how reconnaissance and scanning work. Write the IPv6 security policy separately and test the controls.
Pitfall 5: Poor Address Planning
The vast address space tempts casual allocation, and a poor hierarchy means renumbering later. Plan the hierarchy around your routing topology with the IPv6 subnet calculator before the first prefix is assigned.
Pitfall 6: Over-Relying on Translation
NAT64/DNS64 makes IPv6-only deployment possible, but a translator in every path becomes a bottleneck and a single point of failure. Use translation for specific use cases, not as a blanket solution.
Cost-Benefit Analysis Framework
IPv4 Purchase vs. Lease
Purchase Makes Sense When:
- Need is permanent (about seven years or more at the August 2026 spreads)
- Capital budget available
- Total cost over the holding period is lower than the lease
Leasing Makes Sense When:
- Need is temporary or uncertain
- Capital budget limited
- OpEx treatment preferred
- Flexibility matters more than holding the registration
- IPv6 transition planned within 5 years
IPv6 Migration ROI
Investment side: equipment upgrades where the current gear cannot do IPv6, planning and design, training, and a higher operating cost while both protocols run.
Return side: fewer IPv4 purchases and leases for new capacity, one less stateful hop where carrier-grade NAT goes away, and a simpler addressing plan once the transition is over. The one anchor with a public price is the cloud: on AWS every public IPv4 address costs $0.005 per hour since 1 February 2024, about $43.80 a year, so 1,024 addresses cost about $44,850 a year. That is the line an IPv6-first design reduces.
Via-Registry.com Services for Your IP Strategy
We act as a sponsoring LIR in the RIPE NCC region (Virtua Systems SAS, France), so everything below applies to networks with an active presence there. A leased IPv6 /48 costs €40 per year and a /44 €80 (as of September 2026), larger sizes on request. It is PA space from our allocation: we create its ROA and reverse DNS, and it returns to the pool when the lease ends. An IPv6 PI /48 is assigned by the RIPE NCC to your own organization and stays with you if you ever leave us. It costs €300 setup plus €135 per year, €435 the first year, with RPKI through your own End User hosted CA.
To announce either block from your own network you need an ASN: ASN registration through our sponsorship costs €130 per year, individuals included. IPv4 leases are quoted after verification (from €0.50 per IP per month). If you already hold an IPv4 PI block or an ASN and your LIR is closing, you can change your sponsoring LIR and keep the resource.
Conclusion: Building Your IP Strategy
IPv4 and IPv6 will coexist longer than most roadmaps assume. The real work lies in sequencing the transition: dual-stack remains the safer default for existing infrastructure, while IPv6-primary suits greenfield and cloud-native deployments where NAT64/DNS64 can absorb the remaining IPv4 calls. Treat IPv4 as the strategic resource it has become. Audit utilization before buying or leasing more, and set a fixed timeline for reducing dependence rather than leaving it open-ended.
The organizations that handle this well share a pattern: they started IPv6 planning early, secured budget and executive commitment before the pressure hit, moved in phases, and kept tracking traffic ratios and IPv4 dependencies along the way.
Start with the IPv4 audit in Step 1. It tells you how many addresses you really need to buy, lease or give back, and that number sets the budget for everything else.
FAQ
Is IPv6 faster than IPv4?
Not by itself. Performance depends on the path, and dual-stack clients following RFC 8305 try IPv6 first and fall back to IPv4 quickly, so a broken IPv6 path costs a short delay rather than a failed connection. IPv6 gains where it removes carrier-grade NAT from the path.
How much does an IPv4 address cost in 2026?
About $23.50 per address for /22 to /24 blocks and about $14 for /16 and larger in August 2026, according to broker IPv4.Global's monthly averages. Leasing on the IPXO platform averaged $0.35 per address per month over the six months to September 2026. Registry transfer fees come on top at ARIN and APNIC; RIPE NCC transfers are free.
Can I still get IPv4 from the RIPE NCC?
Only through the waiting list, and only once: each LIR can receive a single /24 and never more (ripe-826). On 25 September 2026, 791 LIRs were queued and the first had waited 485 days. Everyone else buys on the transfer market or leases.
Official References
- Google IPv6 Adoption Statistics: per-country and global IPv6 user share
- APNIC Labs IPv6 Measurement Maps: alternative methodology, 30-day rolling averages
- Internet Society Pulse: multi-source IPv6 deployment tracker
- RIPE NCC IPv4 waiting list and ARIN IPv4 waiting list: live queue lengths
- NRO Internet Number Resource Reports: quarterly RIR statistics
- IPv4.Global sales reports and IPXO market stats: live transfer market indices