Skip to content
Basics

What Are Layer 2 Rollups? Optimistic vs ZK Explained

Layer 2 rollups scale Ethereum by processing transactions off-chain and posting proofs back to the base layer. Here is how optimistic and zero-knowledge rollups differ.

Sofia Lindqvist

Explainers Lead · Jun 9, 2026 · 8 min read

LAYER-2

Why does Ethereum need Layer 2 at all?

Ethereum's base chain, the Layer 1 or L1, can only process a limited number of transactions per second because every node must independently verify every transaction. When demand rises, users bid up gas fees to compete for scarce block space, and simple actions can become expensive. Rather than force all activity onto one congested layer, the ecosystem moved computation to a second tier.

A Layer 2, or L2, is a separate network that runs on top of Ethereum. It executes transactions in its own environment, bundles many of them together, and periodically posts a compressed record back to L1. Users get faster and cheaper transactions, while still inheriting the security of Ethereum because the final record lives on the base chain. The dominant L2 design today is the rollup.

How does a rollup actually work?

A rollup does exactly what the name suggests: it rolls up hundreds or thousands of transactions into a single batch, then submits that batch to Ethereum. The heavy computation happens off-chain on the L2, but the transaction data and a summary of the results are published to L1. This split is the source of the savings.

  • Execution moves off-chain, so you are not paying L1 gas for every individual step.
  • Data and proofs go on-chain, so anyone can reconstruct or verify the L2's state from Ethereum.
  • Costs are shared, because the fixed cost of posting a batch is spread across everyone in it.

The critical question for any rollup is how Ethereum knows the batched results are honest. That is where the two main families diverge: optimistic rollups and zero-knowledge rollups.

What is the difference between optimistic and ZK rollups?

An optimistic rollup assumes every batch is valid by default, hence the name. It posts the results without a mathematical proof, then opens a challenge window, typically around seven days, during which anyone can submit a fraud proof showing a batch was invalid. If a valid challenge succeeds, the fraudulent batch is reverted and the dishonest operator is penalised. The trade-off is that withdrawing funds back to L1 can require waiting out the challenge period, unless you use a third-party liquidity provider to exit faster.

A zero-knowledge rollup, or ZK-rollup, takes the opposite approach. It generates a validity proof, a cryptographic guarantee that the batch was computed correctly, and posts that proof alongside the data. Ethereum verifies the proof mathematically, so there is no need to trust the operator and no lengthy challenge window. Withdrawals can finalise much faster. The cost is computational: producing these proofs is technically demanding, though the technology has matured quickly.

Optimistic rollups trust first and verify only if challenged. ZK-rollups verify upfront with maths. Both post data to Ethereum, so both stay anchored to L1 security.

What should you watch when using an L2?

L2s are not all equally decentralised, and the differences affect real risk. A few things worth checking before moving significant value:

  • The sequencer — the component that orders and batches transactions. Many L2s still run a single centralised sequencer, which cannot steal your funds but could censor or delay transactions. Teams generally publish roadmaps toward decentralising it.
  • Withdrawal times — optimistic rollups have that challenge-period delay for native withdrawals; plan around it or use a bridge with liquidity.
  • Data availability — confirm the rollup posts its data to Ethereum rather than to a separate, less secure system, because on-chain data is what lets anyone reconstruct the state.
  • Gas token — most L2s still charge fees in ETH, and those fees are lower partly because Ethereum introduced cheaper data space, sometimes called blobs, specifically for rollups.

The broader picture is that Ethereum has adopted a rollup-centric roadmap: keep the base layer lean and secure, and let a competitive market of L2s handle throughput. For users, that means most day-to-day activity increasingly happens on an L2, with the L1 acting as a settlement and security backstop. Knowing whether you are on an optimistic or ZK rollup, and how its sequencer and withdrawals work, tells you what you are actually trusting.

Share
Sofia Lindqvist

Explainers Lead

Sofia turns dense on-chain mechanics into plain English. She writes Coin Currents Daily's Learn desk and edits the glossary.