ENSv2 Scraps Its Own L2 and Bets Everything on Ethereum — Here's Why That Matters

In February 2026, Ethereum Name Service did something almost no crypto project has ever done: it killed its own Layer 2 blockchain. After months of building Namechain — a dedicated ZK rollup designed to house the next generation of ENS infrastructure — the team pulled the plug and announced that ENSv2 would deploy exclusively on Ethereum mainnet. The reason? Ethereum's L1 had already solved the problem Namechain was designed to fix.

This decision didn't just reshape ENS's technical roadmap. It sent a signal that reverberates across the entire L2 ecosystem: the rollup-centric future Ethereum once promised may be far smaller than anyone imagined.

The 99% Gas Collapse That Changed Everything

A year ago, registering an ENS name cost roughly $5 in gas fees. Today, it costs less than five cents. That 99% reduction didn't come from some clever L2 trick — it came from Ethereum itself.

Throughout 2025, Ethereum's gas limit doubled from 30 million to 60 million units. Combined with continued optimization of EIP-4844 blob data and overall network efficiency improvements, the base layer became dramatically cheaper. For ENS, this eliminated the core economic argument for building a separate chain.

"The math simply changed," ENS Labs explained in their February 2026 announcement. With registration costs already at sub-nickel levels on mainnet, the overhead of maintaining a dedicated rollup — sequencer infrastructure, bridge security, cross-chain complexity — no longer made sense.

The decision freed up engineering resources. ENS reported that 80% of development effort could now focus on ENSv2's core features rather than maintaining custom blockchain infrastructure. For a protocol that serves as Web3's identity layer, that focus matters.

What ENSv2 Actually Brings

Stripping away the L2 distraction reveals what ENSv2 is really about: a ground-up architectural rewrite of how naming works on Ethereum.

Hierarchical Registries. Each name can now provide its own registry implementation for subnames. This means name owners and developers get direct control over ownership and transfer rules. A DAO could implement governance-gated subname distribution. A company could enforce compliance rules on its subdomain hierarchy. The flexibility is structural, not cosmetic.

Programmable Name Records. ENSv2 introduces a new ownership model with improved expiration handling and flexible record management. Names become programmable primitives, not static pointers. A single .eth name can resolve to addresses across more than 100 different cryptocurrency networks — Bitcoin, Solana, Ethereum L2s, and dozens more — simultaneously.

Streamlined Registration. The ENS App and ENS Explorer public alphas, already available for testing, demonstrate a dramatically simplified registration flow. Multi-chain support, flexible ownership models, and name management are handled through an interface that abstracts away cross-chain complexity.

Cross-Chain Resolution Without Cross-Chain Infrastructure. Here's the counterintuitive part: by staying on L1, ENS actually improved its L2 interoperability. ENSv2's architecture enhances resolution across all existing L2s — Base, Arbitrum, Optimism, and others — without requiring users to bridge to a dedicated ENS chain. The protocol resolves names wherever users already are.

2.8 Million Names and Growing

With over 2.8 million .eth domains registered and registrations growing at 8% month-over-month, ENS has quietly become one of the most adopted protocols in crypto. These aren't speculative registrations driven by airdrop farming — they represent persistent demand for human-readable Web3 identity.

The 99% gas cost reduction is expected to accelerate this growth further. At sub-five-cent registration costs, the economic barrier to claiming a .eth name essentially disappears. For context, traditional DNS domain registration on services like GoDaddy typically costs $10-15 per year. ENS is now cheaper than Web2 naming.

But raw registration numbers tell only part of the story. The more significant metric is integration depth. ENS names are now supported by every major wallet (MetaMask, Coinbase Wallet, Rainbow, Phantom), most prominent DApps, and an expanding set of traditional platforms. When you send crypto to "alice.eth" instead of "0x7f3a...b2c1," ENS is doing the resolution.

The Vitalik Factor: L1 Scaling Changes the L2 Calculus

ENS's decision didn't happen in isolation. It reflects a broader philosophical shift championed by Ethereum co-founder Vitalik Buterin himself.

Throughout late 2025 and into 2026, Buterin increasingly questioned the rollup-centric roadmap he had previously championed. His core observation: Ethereum's base layer was scaling faster than anyone had anticipated, while L2s had struggled to meaningfully decentralize.

Many rollups still relied on centralized sequencers, had limited fraud or validity proof implementations, and fragmented liquidity across isolated ecosystems.

The ENS decision validated this critique with a concrete example. If one of Ethereum's most important infrastructure protocols concluded that L1 scaling made a dedicated L2 unnecessary, what does that say about the dozens of application-specific rollups launched in 2024-2025?

The implications extend beyond ENS. The L2 landscape is now expected to consolidate into two categories: commodity Ethereum-equivalent rollups competing on fees and throughput, and specialized chains with fundamentally different execution models (like Solana Virtual Machine rollups or privacy-focused environments). The middle ground — application-specific L2s that could have run on a sufficiently scaled L1 — is shrinking.

ENS as the Identity Layer for AI Agents

Perhaps ENSv2's most forward-looking impact is its intersection with the emerging AI agent economy.

ERC-8004, which went live on Ethereum mainnet in January 2026, establishes on-chain identity, reputation, and validation registries for autonomous AI agents. Developed by contributors from MetaMask, the Ethereum Foundation, Google, and Coinbase, the standard treats ENS names as first-class identifiers.

Under ERC-8004, an AI agent's ENS name functions as its human-readable handle — resolved before evaluating reputation or validation data. Each agent is registered as an NFT, providing a portable, censorship-resistant, and globally unique on-chain identity.

The protocol links blockchain registries to flexible configuration files where endpoints can be added dynamically, combining AI primitives (Model Context Protocol, Agent-to-Agent) with Web3 primitives (wallet addresses, DIDs, and ENS names).

This creates a practical architecture where ENS names serve triple duty: they identify the agent to humans, provide machine-readable resolution for other agents, and anchor reputation data on-chain. As projections suggest 60% of crypto wallets may incorporate agentic AI capabilities by late 2026, ENS's position as the universal naming layer becomes increasingly strategic.

The trust models are designed to be proportional to risk. Low-stakes tasks (like information retrieval) might rely on simple reputation scores. High-value transactions could require stake-secured re-execution, zero-knowledge proofs, or Trusted Execution Environment attestation — all anchored to an agent's ENS identity.

What This Means for Ethereum's Future

ENS's pivot carries implications that extend well beyond naming infrastructure.

For L2 builders: The era of "we need our own chain" is ending. Unless a project genuinely requires a different execution environment or has governance reasons for chain sovereignty, building on Ethereum L1 — or deploying on existing general-purpose L2s — is increasingly the rational choice. The ENS team's frank assessment that maintaining custom blockchain infrastructure was a distraction rather than an enabler should give pause to every team planning an application-specific rollup.

For Ethereum itself: ENS's decision is validation. The gas limit increase, blob data optimizations, and upcoming Pectra upgrade demonstrate that L1 scaling is not a distant promise but a present reality. Ethereum's roadmap through 2029 — including slot time reductions, validator set restructuring, and native data availability sampling — suggests the base layer will continue to absorb use cases previously assumed to require L2s.

For Web3 identity: With ENSv2's programmable name records, cross-chain resolution, and AI agent integration, .eth names are evolving from vanity addresses into foundational infrastructure. They serve as the connective tissue between human users, autonomous agents, and multi-chain ecosystems — all anchored to Ethereum's security guarantees.

ENS made the rare decision to build less infrastructure, and ended up with a more powerful product. In an industry addicted to new chains, new layers, and new complexity, that restraint might be its most radical innovation.

BlockEden.xyz provides enterprise-grade Ethereum RPC and API infrastructure for developers building on the protocols and identity systems that define Web3. Explore our API marketplace to power your next project with reliable, low-latency access to Ethereum and its growing ecosystem.