The End of the L2 Era? Vitalik Buterin Reverses Stance on Ethereum Scaling

Ethereum co-founder Vitalik Buterin has delivered one of the most consequential philosophical shifts in the network’s history, publicly reversing his long-held belief that layer-2 networks should be Ethereum’s primary path to scale.

In a post shared on X on Tuesday, Buterin said the original vision for layer-2s “no longer makes sense,” arguing that many have failed to decentralize and that Ethereum’s base layer is now capable of scaling more directly through protocol-level upgrades.

“We need a new path,” Buterin wrote, pointing to two parallel developments: the Ethereum mainnet’s increasing capacity through gas-limit increases, and the coming integration of native rollups. Taken together, he said, these changes undermine the premise that Ethereum must rely on external execution layers to achieve meaningful throughput gains.

Layer-2s were originally conceived as extensions of Ethereum, designed to handle most transactions cheaply and at high speed while inheriting Ethereum’s security guarantees. In theory, this meant transaction validity, censorship resistance, and finality would ultimately be enforced by the mainnet.

Buterin now argues that reality has diverged sharply from that vision.

“If you create a 10,000 TPS EVM where its connection to L1 is mediated by a multisig bridge, then you are not scaling Ethereum,” he said, criticizing designs that rely on trusted intermediaries rather than cryptographic guarantees.

Layer-2s Should Find Their Own Purpose

The comments represent a notable reframing of Ethereum’s technical roadmap, which for years placed layer-2s at the center of scaling efforts. 

Networks such as Arbitrum, Optimism, Base, and Starknet have collectively attracted billions of dollars in value and millions of users.

Rather than dismissing those networks outright, Buterin suggested they pivot away from being framed as Ethereum’s scaling engine and instead specialize in distinct application niches. He highlighted areas such as privacy, identity, finance, social applications, and artificial intelligence as domains where independent execution layers could still provide meaningful innovation.

That shift would mark a conceptual break from the idea that Ethereum’s future throughput must be outsourced.

Mainnet Scaling Back in Focus

Buterin’s remarks also align with a growing faction of Ethereum developers who have argued that the base layer itself deserves renewed attention. Among them is Max Resnick, a former researcher at Consensys, who later moved to the Solana ecosystem after failing to rally enough support for an L1-first scaling strategy within Ethereum.

The shift has now received public backing from prominent ecosystem voices. Ryan Sean Adams, co-host of the Bankless show, called Buterin’s comments “the pivot,” adding: “Strong ETH, Strong L1.”

Native Rollups and Gas Limits Take Center Stage

At the core of Buterin’s revised outlook is the growing role of native rollups—execution environments that are built directly into Ethereum rather than layered on top of it. Unlike traditional rollups, which bundle transactions off-chain and post data back to Ethereum, native rollups are verified directly by Ethereum validators.

Buterin said this architecture becomes especially powerful once zero-knowledge Ethereum Virtual Machine proofs are integrated at the protocol level, allowing Ethereum to scale while preserving its trust-minimized security model.

In parallel, developers have already begun pushing the network’s capacity higher. 

In mid-December, Ethereum contributors discussed raising the gas limit from 60 million to 80 million following the implementation of the second blob-parameter-only hard fork, which went live in January. Increasing the gas limit allows more transactions and smart-contract operations per block, potentially improving throughput while easing fee pressure.

Looking further ahead, Ethereum researcher Justin Drake has outlined a decade-long roadmap that targets 10,000 transactions per second on the Ethereum mainnet—up dramatically from the roughly 15 to 30 TPS seen today.

If realized, that vision would fundamentally reshape Ethereum’s scaling narrative, placing the base layer back at the center of its own growth story and forcing layer-2 networks to redefine their role in the ecosystem.