CoinucationWhat is a layer 2?
7 min read · Updated Oct 6, 2026 · By Coinucation Editorial
- A layer 2 processes transactions off the main chain and settles results back to it, inheriting the base chain's security.
- Rollups bundle transactions and post compressed data to Ethereum, which lets anyone reconstruct their state.
- Optimistic rollups rely on fraud proofs and a withdrawal window; ZK rollups prove validity cryptographically.
- Centralized sequencers, upgrade keys, and bridge risk are the main concerns with current layer 2s.
The short answer
A layer 2, or L2, is a separate network built on top of a base blockchain, called the layer 1, that handles transactions more cheaply and quickly while relying on the base chain for security and final settlement. The idea is to move most of the activity off the crowded main chain, then periodically post a compressed summary back to it.
The term is most often used in the context of Ethereum, where networks like Arbitrum, Optimism, Base, and zkSync process the majority of everyday user transactions. Fees on these networks are usually a few cents, compared with dollars on Ethereum itself during busy periods. Bitcoin also has layer 2 systems, most notably the Lightning Network for fast small payments.
Layer 2s exist because base blockchains deliberately limit how many transactions they process in order to stay decentralized. Rather than loosen those limits, the layer 2 approach keeps the base chain secure and conservative while letting activity scale on top of it. The result is a layered system where security lives at the bottom and most activity happens above it.
How rollups work
The dominant layer 2 design on Ethereum is the rollup. A rollup runs its own execution environment where users send transactions and interact with applications, just as they would on Ethereum. Instead of each transaction being recorded on Ethereum individually, the rollup bundles hundreds or thousands together, compresses them, and posts the batch to the main chain as a single piece of data.
Because the data is posted to Ethereum, anyone can reconstruct the rollup's state from the base chain alone. This is what ties the layer 2's security to the layer 1. Even if every operator of the rollup disappeared, users could in principle recover their funds from the data on Ethereum. Rollups also inherit Ethereum's ordering and censorship resistance for the batches they post.
A March 2024 Ethereum upgrade called Dencun introduced a dedicated cheap data space for rollups, known as blobs. This dramatically reduced the cost of posting batches, and layer 2 fees fell sharply as a result. The cost a user pays on a rollup is mostly their share of that posting cost plus a small fee for the rollup's own execution.
Optimistic rollups versus ZK rollups
There are two main ways a rollup proves to Ethereum that its batches are correct. Optimistic rollups, such as Arbitrum, Optimism, and Base, assume batches are valid by default and give a window, typically about seven days, during which anyone can submit a fraud proof showing a batch was wrong. If a fraud proof succeeds, the bad batch is reverted and the dishonest party is penalized.
Zero knowledge rollups, or ZK rollups, such as zkSync, Starknet, Scroll, and Linea, take the opposite approach. They generate a cryptographic proof for every batch demonstrating that all transactions were executed correctly. Ethereum verifies the proof, which is fast, and the batch is final immediately. No waiting period is needed because validity is proven rather than assumed.
The practical difference shows up when withdrawing to Ethereum. From an optimistic rollup, a standard withdrawal takes the full challenge window, though third party bridges offer faster exits for a fee. From a ZK rollup, withdrawals complete once the proof is verified, typically within hours. ZK technology is more complex to build but is widely regarded as the long term direction.
- Optimistic rollups: assume validity, allow fraud proofs, roughly seven day withdrawal window
- ZK rollups: prove validity with cryptography, faster withdrawals, harder to build
- Both post data to Ethereum so state can be reconstructed from the base chain
Other layer 2 approaches
Bitcoin's Lightning Network takes a different approach called payment channels. Two parties lock funds in a shared on chain transaction, then exchange signed updates off chain as many times as they like. Only the opening and closing transactions touch the Bitcoin blockchain. Payments can route across a network of channels, enabling near instant, very cheap transfers.
Sidechains are independent blockchains with their own validators that connect to a main chain through a bridge. Polygon's original proof of stake chain is a well known example. Sidechains can be fast and cheap, but because they do not post their data or proofs to the base chain, their security depends on their own validators rather than on the layer 1. Many people do not consider them true layer 2s for this reason.
Validiums and similar designs use validity proofs like a ZK rollup but store the data off chain to cut costs further. This makes them cheaper but introduces a dependency on whoever holds the data. The spectrum from full rollup to sidechain is really a spectrum of how much security is inherited from the base chain versus provided independently.
Risks and tradeoffs
Most layer 2s today are not as decentralized as the chains they settle to. Many run a single sequencer, the component that orders transactions, operated by the founding team. If it goes down, the network can pause. Many also have upgrade keys controlled by a small group, which means the code could in principle be changed. Projects publish roadmaps toward decentralizing these components, but progress varies.
Bridges between layers are a historical weak point. Moving assets between a layer 1 and a layer 2, or between two layer 2s, often involves a bridge contract, and bridge exploits have accounted for some of the largest thefts in crypto history. Using the official, canonical bridge for a network is generally safer than third party alternatives, though slower.
Fragmentation is another cost. With dozens of layer 2s, liquidity and users are spread across many networks that cannot talk to each other directly. Moving between them costs time and fees, and applications may exist on one but not another. Various interoperability efforts aim to reduce this friction, but it remains a real inconvenience.
How to get started
To use an Ethereum layer 2, add the network to a compatible wallet such as MetaMask, then move a small amount of ether to it. The most reliable route is the network's official bridge, though many exchanges now allow direct withdrawals to major layer 2s, which is usually cheaper and faster. Once funded, you can use applications exactly as you would on Ethereum, paying gas in ether.
Start with one of the large, established networks and a small amount. Look at how long the network has operated, how much value it holds, who controls its sequencer and upgrade keys, and whether it posts full data to Ethereum. Independent sites track these decentralization properties for major rollups. A network's own documentation usually states its current stage of decentralization honestly.
Keep in mind that each layer 2 is its own environment. Tokens on Arbitrum are not automatically on Base, and an address funded on one network has a zero balance on another until you bridge. Checking which network your wallet is on before every transaction avoids most mistakes. Sending tokens to the right address on the wrong network is a frequent and costly error.