Buterin says L2-first plan is over, shift scaling to mainnet

Ethereum co-founder Vitalik Buterin wrote on X on Tuesday that the network’s long-time plan to scale mainly through external layer-2 networks no longer adds up. He argued that future capacity should come from the base chain via native rollups and by increasing the gas limit.

He contended many L2s have not inherited Ethereum’s security and decentralization as planned, which undermines the idea of moving most activity off-chain. He noted the mainnet can handle more throughput today and will gain further headroom from protocol changes.

“The original vision of L2s and their role in Ethereum no longer makes sense,” Buterin wrote, calling for a reordering of priorities. He added: “If you create a 10000 TPS EVM where its connection to L1 is mediated by a multisig bridge, then you are not scaling Ethereum.” He also stated, “We need a new path.”

Rather than serve chiefly as throughput engines, he suggested L2s focus on specific niches, including privacy, identity, finance, social applications and AI.

Buterin expressed growing confidence in native rollups-precompiled systems embedded in Ethereum so validators verify transactions directly-especially once zero-knowledge Ethereum Virtual Machine proofs are integrated at the base layer. By contrast, today’s rollups batch and execute transactions off-chain before posting data back to Ethereum.

He also referenced active discussions among Ethereum developers about lifting the gas limit, which defines how much computation fits in each block. In December, teams weighed moving from 60 million to 80 million after a blob-parameter-only hard fork slated for January, which has since taken effect. A higher cap would let more transactions and contract calls fit in each block and could lower fees.

His remarks land amid an internal debate about where scaling should occur. Some contributors have pushed for a mainnet-first path. Former Consensys researcher Max Resnick advanced that view and later shifted his work to the Solana ecosystem.

Ethereum’s roadmap has long highlighted L2s as the primary scaling path. The base chain processes roughly 15 to 30 transactions per second today. In July, researcher Justin Drake sketched a 10-year plan to reach about 10,000 transactions per second on the mainnet through data availability upgrades and advanced proof systems.

As we reported earlier, on Jan. 18, 2026 Vitalik Buterin urged core developers to add an explicit “garbage collection” phase to Ethereum’s roadmap, saying years of additive upgrades and strict backward-compatibility have bloated the code and weakened independent verification. 

He called for simplicity, fewer consensus-critical cryptographic primitives, stronger invariants, demoting rarely used features to contracts, and gas costs aligned to real resource use, citing the proof-of-stake transition and recent gas-cost reforms as past cleanups.