Ethereum co‑founder Vitalik Buterin has published a new post arguing that the network’s original vision for Layer‑2s (L2s) as “branded shards” of Ethereum no longer makes sense. He points to two realities shaping the ecosystem in 2026: L2 progress toward stage‑2 decentralization has been slower than expected, and Ethereum’s base layer (L1) itself is scaling with low fees and projected gas limit increases.
Shifting Vision
Buterin explained that Ethereum scaling was originally defined as creating large amounts of block space backed directly by Ethereum’s security guarantees. Under that definition, rollups connected by multisig bridges were never true scaling. Now, with L1 capacity expanding, he argues Ethereum does not need L2s to serve as shards. Instead, L2s should be seen as a spectrum of chains with varying degrees of connection to Ethereum, each offering unique features.
What L2s Should Do
Buterin suggested that L2 projects identify value beyond scaling. Examples include privacy‑focused virtual machines, ultra‑low‑latency sequencing, specialized application efficiency, or non‑financial use cases such as identity and AI. He emphasized that L2s handling ETH or Ethereum‑issued assets should at least meet stage‑1 standards, and maximize interoperability with Ethereum even if they diverge from the EVM model.
Native Rollup Precompile
On Ethereum’s side, Buterin highlighted the potential of a native rollup precompile that verifies ZK‑EVM proofs directly within Ethereum. This would provide trustless interoperability, synchronous composability, and security‑council‑free verification. He argued such a design would allow L2s to extend Ethereum safely while adding their own innovations.
Outlook
Buterin’s remarks mark a pivot in Ethereum’s scaling narrative: L1 expansion reduces reliance on L2s for throughput, while L2s are encouraged to differentiate with specialized features. The debate underscores Ethereum’s evolving roadmap as it balances decentralization, scalability, and institutional adoption.
