Points of Focus

• Ethereum’s Layer 1 scaling weakens the original case for Layer 2s as “branded shards.”
• Vitalik Buterin argues that many L2s no longer meet the definition of scaling Ethereum.
• The next phase shifts L2s from cost reducers to differentiated systems with clear guarantees.

Ethereum co-founder Vitalik Buterin is pressing the ecosystem to revisit a core assumption that has guided development for years. As Ethereum’s base layer scales faster than expected, he says the long-standing rollup-centric model no longer fits reality, a claim grounded in two measurable shifts that now shape how you use Ethereum.

Why the rollup-centric roadmap is under pressure

Buterin points to slow progress among L2s toward higher decentralization stages, with many still at stage 1 and some signaling they may never move further due to regulatory or control requirements. At the same time, Ethereum itself is scaling, with fees remaining low and gas limits projected to rise through 2026, a shift Buterin outlined in a February 2025 post on L1 scaling where he argued that expanding base-layer capacity remains essential even as rollups mature.

In a recent post on X, Buterin argued that these trends break the original vision that placed L2s at the center of Ethereum scaling. That model assumed L2s would inherit Ethereum’s security while handling most transactions. If Ethereum can now provide cheap block space directly, that tradeoff looks different.

There have recently been some discussions on the ongoing role of L2s in the Ethereum ecosystem, especially in the face of two facts:

* L2s' progress to stage 2 (and, secondarily, on interop) has been far slower and more difficult than originally expected
* L1 itself is scaling,…

— vitalik.eth (@VitalikButerin) February 3, 2026

What “scaling Ethereum” actually means

Buterin draws a firm line around definitions. Scaling Ethereum means creating block space backed by Ethereum’s full security guarantees. Systems that rely on multisig bridges or retain override control do not meet that bar, even if throughput is high. His point is blunt: a fast chain that can be paused or altered by a small group is not extending Ethereum’s trust model.

This matters for how projects describe themselves. Some L2 teams openly accept tradeoffs to meet compliance or customer demands. That may suit their users, but it does not, in Buterin’s framing, count as scaling Ethereum. The mismatch is semantic but also strategic. Mislabeling clouds user expectations and risk.

A spectrum, not a hierarchy, for Layer 2s

Rather than pushing every L2 toward the same endpoint, Buterin proposes treating them as a spectrum. Some may stay tightly coupled to Ethereum’s security, while others optimize for different goals. Privacy-focused virtual machines, application-specific execution, non-financial systems like identity or social, or extreme performance beyond what L1 can offer are all valid paths.

The key is clarity: if you use an L2, you should know which guarantees it provides and which it does not. Security stage, bridge design, and control assumptions need to be explicit, not implied.

Ethereum’s side of the bargain

Buterin also points to protocol work that could tighten the link where it matters. A native rollup precompile that verifies ZK-EVM proofs inside Ethereum would lower reliance on security councils and custom bridges. If implemented, it would let L2s prove EVM execution natively while extending their own features on top, improving interoperability and opening the door to synchronous composability where actions across chains settle together.

The message to builders is direct: do not position your project as a thinner version of Ethereum, but add something Ethereum does not. As Layer 1 capacity grows, the value of L2s will rest less on cheaper fees and more on distinct capabilities you cannot get on mainnet alone.