Vitalik Buterin Defends Centralized Sequencers Amid Calls for More Decentralized L2s

Key Takeaways

• Vitalik Buterin recently discussed the advantages of centralized L2 sequencers.
• The risks of centralized sequencing include censorship and central points of failure.
• Emerging alternatives include based sequencers and decentralized sequencers.

Proponents of decentralization have often criticized popular Ethereum Layer 2s, like Base, Arbitrum and Polygon, for their reliance on centralized sequencers.

However, in a recent AMA on Tako, the Ethereum (ETH) founder insisted, “Centralized sequencers actually have many advantages.”

What Are Centralized Sequencers?

Unlike layer-1 blockchains that share responsibility for ordering and batching transactions among multiple block proposers, L2s typically deploy a single sequencer each.

This reliance on a single machine, usually operated by the L2s governing entity or foundation, creates a single point of failure and poses the risk of censorship.

Vitalik Buterin defended L2s’ use of centralized sequencers and pointed to three advantages of the standard setup.

Buterin Weighs In

Firstly, he said centralized sequencers can mitigate frontrunning, a problem that has plagued the Ethereum Layer 1.

Because they have a single point of control over the transaction queue, L2s can implement strategies that reduce the opportunities for malicious actors to observe pending transactions and exploit them.

Another advantage of centralized sequencers, Buterin pointed out, is that they can deliver “instant preconfirmations.”

In other words, since sequencers organize transactions off-chain, they aren’t subject to the latency of on-chain consensus or block propagation delays. They can, therefore, instantly accept transactions into the queue.

The final advantage of the traditional L2 sequencing model highlighted by Buterin is more practical, “It is easy to turn a traditional application into a blockchain application because the server directly becomes a sequencer.”

Centralized Sequencing Risks

Despite his defense of standard L2 sequencing, Buterin acknowledged that centralized sequencers still have risks.

Sequencer operators, whether private companies, non-profit foundations or decentralized committees, essentially have the power to shut L2s down and suspend all transactions.

This happened in 2024 when Linea, facing the threat of hackers, made the decision to suspend its sequencer and “censor attacker addresses.”

While their intentions were honorable, the incident highlighted how much power a small team of developers had over the L2.

In other cases, sequencer downtime has resulted from technical challenges rather than orchestrated shutdowns. For instance, dYdX’s L2 experienced a major outage in 2024 that left users of the exchange unable to trade for hours.

Alternatives to Centralized Sequencing

While the biggest L2s have mostly stuck to centralized sequencers, there are a growing number of smaller platforms exploring alternatives.

In 2022, Metis was launched as the first L2 that let multiple participants take on sequencing roles. Another project, Rome Protocol, uses the high-throughput Solana (SOL) chain as a shared sequencer for multiple L2s.

Finally, Syndicate’s on-chain sequencing network uses a technique called “metabased” sequencing.

Like based sequencing, this approach relies on a separate blockchain to order transactions. However, because the sequencing chain is specially optimized for the task, it doesn’t have the same limitations.

Metabased sequencing is especially useful for cross-chain applications, because it can sequence transactions from multiple L2s. And further down the line, existing L2s may also adopt the technology.