Ethereum processes only a limited number of transactions per second on its main chain. During busy periods, that ceiling shows — fees spike and confirmations slow to a crawl. Layer 2 rollups exist to fix that without rebuilding the base layer from scratch.
The core idea is straightforward: instead of settling every single transaction directly on Ethereum's main chain (called Layer 1, or L1), rollups bundle hundreds or thousands of transactions together, process them somewhere else, and then post a compressed summary back to L1. The main chain still acts as the final judge — it holds the security and the ultimate record — but most of the computational work happens off it.
Two competing designs dominate the rollup landscape: optimistic rollups and zero-knowledge rollups (commonly called ZK rollups). They both achieve the same goal of moving computation off-chain, but they take fundamentally different approaches to one critical question: how do you prove the off-chain work was done honestly?
**Optimistic Rollups: Assume First, Challenge Later**
Optimistic rollups work on a presumption of innocence. When a batch of transactions is submitted to Layer 1, the system assumes by default that everything in the batch is valid. It does not immediately verify every transaction. Instead, it posts the data and opens a window — typically several days — during which anyone watching the network can raise a challenge if they spot fraud.
If a challenger finds something wrong, they submit what is called a "fraud proof." This triggers a process where the disputed computation is re-run on L1 to check who is right. If fraud is confirmed, the bad actor loses a financial deposit and the incorrect state is rejected. If nobody raises a challenge within the window, the batch is considered final.
The trade-off is time. Because of that challenge window, withdrawing funds from an optimistic rollup back to L1 can take up to a week. Users who need faster exits can use third-party "liquidity providers" who front the funds immediately for a fee, but that adds a layer of complexity and cost. The upside is that optimistic rollups are relatively straightforward to build and are compatible with the Ethereum Virtual Machine (EVM), meaning developers can port existing Ethereum smart contracts over with minimal changes.
Arbitrum and Optimism are two of the most widely used optimistic rollup networks.
**Zero-Knowledge Rollups: Prove It Up Front**
ZK rollups take the opposite approach. Rather than assuming correctness and waiting for a challenge, they generate a cryptographic proof that the entire batch of transactions was computed correctly — before posting anything to L1. This proof, called a validity proof (or sometimes a ZK-SNARK or ZK-STARK depending on the system), is a compact piece of math that L1 can verify almost instantly.
Think of it like a student showing their work on an exam versus a student handing in only answers. The optimistic approach accepts the answers and waits to see if a teacher contests them. The ZK approach attaches an automatically checkable certificate proving the work was done correctly.
Because validity is proven before anything is accepted on-chain, withdrawals from ZK rollups can be nearly instant once the proof is verified. There is no waiting period. From a security standpoint, validity proofs are considered more robust — fraud cannot slip through even temporarily.
The historical challenge with ZK rollups has been complexity. Generating cryptographic proofs is computationally intensive, and building ZK systems that are fully compatible with the EVM has been significantly harder than building optimistic ones. That gap has been narrowing steadily as teams develop more efficient proof systems and EVM-compatible ZK stacks.
zkSync, StarkNet, Polygon zkEVM, and Scroll are among the projects building in this space.
**How Do They Compare?**
The two approaches reflect a classic engineering trade-off between simplicity and mathematical rigor. Optimistic rollups got to market faster and accumulated more developer activity early on because of their EVM compatibility and easier construction. ZK rollups are catching up on compatibility and promise stronger security guarantees and faster finality.
For everyday users, the most visible difference is withdrawal times and, to some extent, transaction fees. For developers, EVM compatibility and the maturity of tooling matter most. For the broader ecosystem, ZK rollups are increasingly seen as the more powerful long-term path because validity proofs scale better and do not rely on the social assumption that someone is always watching for fraud.
**Why This Matters Beyond Ethereum**
The rollup model is not unique to Ethereum, but Ethereum is where it has been developed most thoroughly. Ethereum's own long-term roadmap — sometimes called "the rollup-centric roadmap" — explicitly treats rollups as the primary scaling path, with future upgrades designed to make rollups cheaper and more efficient to use.
Understanding the distinction between optimistic and ZK rollups matters because these two designs are shaping how blockchains scale for millions of users. Each represents a different set of bets about cryptography, developer experience, and trust. Neither is a finished product — both are actively evolving. But together, they represent the most serious attempt yet to make public blockchains fast and affordable without sacrificing the decentralization that makes them useful in the first place.