A Layer 2 improves scalability by moving transaction execution off its base chain, the Layer 1, and posting batched results back to it. The Layer 1 makes the data available and checks or settles disputes over the results instead of running every transaction itself. That lets many more transactions share the Layer 1's security at a lower cost each. Rollups are a widely used Layer 2 design, and they differ mostly in how the Layer 1 learns that a batch is correct.
Last updated: October 5, 2026.
What is a Layer 2?
A Layer 2 is a separate network that executes transactions and relies on a Layer 1 for settlement. A rollup then posts its transaction data and resulting state to the Layer 1, which settles the result once it is proven or no longer open to challenge.
Because a rollup publishes its data to the Layer 1, anyone can use that data to rebuild the rollup's state without trusting its operators.
How do rollups improve scalability?
A rollup gathers many transactions into one batch, compresses the batch, and posts it to the Layer 1 in a single submission. The fixed cost of using the Layer 1 is then shared across every transaction in the batch instead of being paid by each one.
The Layer 1 also does less work per transaction. In normal operation it does not re-execute the batch. It makes the data available and either verifies a proof or gives anyone a window to challenge a wrong result. On Ethereum, a dedicated data space known as blobs was added for this purpose, which lowered the cost of posting rollup data. Blob data stays available for a limited period rather than permanently.
What is the difference between optimistic and zero-knowledge rollups?
The two main rollup types differ in how a batch is shown to be correct.
| Rollup type | How correctness is established | When withdrawals to the Layer 1 finish |
|---|---|---|
| Optimistic | Batches are assumed valid, and anyone can submit a fraud proof during a challenge window | After the challenge window ends |
| Zero-knowledge | Each batch comes with a validity proof that the Layer 1 checks | Once the proof is verified on the Layer 1 |
Optimistic rollups avoid proof costs but make withdrawals wait out the challenge window. Zero-knowledge rollups settle sooner but need heavy computation for each proof.
What are the trade-offs of Layer 2s?
Each Layer 2 is its own environment. Balances and liquidity on one are not automatically available on another, even when both settle to the same Layer 1. Moving assets between them needs a bridge or a cross-chain messaging route. The multichain vs cross-chain overview explains why being deployed on several networks is not the same as connecting them.
Many rollups also rely on a single operator, called a sequencer, to order transactions. Rollup designs typically let users submit a transaction to the Layer 1 directly if the sequencer delays it.
FAQ
What is the difference between Layer 1 and Layer 2?
A Layer 1 runs its own consensus and keeps the final record. A Layer 2 executes transactions separately and posts its data and results to a Layer 1 for settlement.
Do Layer 2s use the same wallet address as their Layer 1?
On EVM-compatible Layer 2s, the same key usually controls the same address, but balances stay separate on each network. The same address on different chains overview covers this.
How do tokens move between Layer 2s?
Through a bridge or a messaging protocol that connects both networks. With cross-chain token standards such as OFT, a transfer debits the sender on the source network and credits the recipient on the destination, so no third party issues a wrapped copy.
Build across Layer 2s
For cross-chain concepts and supported integration paths, start with the LayerZero developer resources.