Layer 1 vs Layer 2: The Evolution of Blockchain Scaling
Layer 1 is the base blockchain, while Layer 2 is built on top to scale it. Learn how both layers work and where blockchain scaling is heading next.
Key takeaways
- Layer 1 is the base blockchain. It runs consensus, validates transactions, and provides the ultimate source of security and settlement.
- Layer 2 is a separate network built on top of a Layer 1, designed to process transactions faster and cheaper while still relying on the base layer for security.
- The two layers are not competitors but a stack: Layer 1 handles trust and finality, Layer 2 handles throughput and user-facing activity.
- The traditional L1/L2 framing is being reshaped by modular architectures, restaking, and app-chains, which split blockchain responsibilities across more layers than just two.
The biggest difference between Layer 1 and Layer 2 is where transactions are processed. Layer 1 is the base blockchain itself, like Bitcoin or Ethereum, where transactions are validated and finalized. Layer 2 is a separate network built on top, designed to handle transactions faster and cheaper before settling the results back to Layer 1.
That distinction sounds simple, but the line between the two has become much harder to draw in 2026 than it was even two years ago. What follows is a closer look at how each layer works, where they overlap, and why the conversation is shifting.
Layer 1 vs Layer 2 Blockchain: Quick Comparison
Layer 1 | Layer 2 | |
| Position in the stack | Base blockchain | Built on top of a Layer 1 |
| Primary role | Consensus, security, final settlement | Transaction execution, throughput |
| Security source | Native (own validators or miners) | Inherited from the underlying Layer 1 (with varying degrees of trust) |
| Throughput | Limited — Bitcoin ~7 TPS, Ethereum ~15 TPS | Significantly higher — often 1,000+ TPS, up to 100,000+ in theory |
| Transaction cost | Higher, more volatile during congestion | Much lower, typically a few cents or less |
| Finality | Native to the chain | Depends on the design (instant for ZK proofs, ~7 days for optimistic withdrawal) |
| Examples | Bitcoin, Ethereum, Solana, Avalanche | Arbitrum, Optimism, Base, zkSync, Lightning Network |
The simplest way to think about Layer 1 and Layer 2 is by their role in the stack: Layer 1 is where security and settlement live, while Layer 2 is where most user activity happens.
Understanding a Layer 1 Blockchain
| In short: A Layer 1 blockchain is the foundational network that processes and finalizes its own transactions without relying on any other chain to function. Everything else in a blockchain ecosystem, including Layer 2 networks, is built on top of it. |
Every Layer 1 runs on three core components:
- Consensus mechanism: how validators or miners agree on the next valid block. Proof-of-Work (Bitcoin) and Proof-of-Stake (Ethereum, Solana, Cardano) are the two dominant families.
- Native security model: validators stake or burn resources to defend the network. There is no other chain backing them up.
- Settlement and state: once a transaction is included in a block and confirmed, it becomes part of the canonical record.
A Layer 1's native token (BTC, ETH, SOL) pays for these services – gas fees, staking, block rewards – and serves as the unit of account inside the network.
Examples of Layer 1 blockchains:
- Bitcoin: the original Layer 1, optimized for security and decentralized settlement, processing roughly 7 transactions per second.
- Ethereum: a smart contract platform handling around 15 TPS on its base layer, intentionally kept simple while pushing scalability to Layer 2.
- Solana: a high-throughput chain using parallel execution to scale on Layer 1 itself, with a theoretical ceiling of 65,000 TPS.
- Avalanche, BNB Chain, Cardano, Sui, Aptos: other major Layer 1s, each making different trade-offs between throughput, decentralization, and programmability.
Pros | Cons |
| ✅ Strongest security guarantees in the stack | ✖ Limited throughput, especially under demand |
| ✅ Maximum decentralization (varies by chain) | ✖ Higher and more volatile gas fees |
| ✅ No external trust assumptions | ✖ Upgrades require coordination (sometimes hard forks) |
| ✅ Native settlement and finality | ✖ Can become congested during major activity spikes |
Understanding a Layer 2 Blockchain
| In short: A Layer 2 blockchain is a secondary network that runs on top of a Layer 1 to handle transactions more efficiently. Instead of processing every transaction directly on the base chain, a Layer 2 batches them off-chain, then periodically posts the results back to Layer 1 for settlement. |
The core idea is to keep the heavy work off the main chain while letting Layer 1 act as the final arbiter. A typical flow looks like this:
- Users submit transactions to the Layer 2 network.
- Layer 2 executes those transactions in its own environment.
- Periodically, Layer 2 posts a compressed batch of transactions (or a proof of their validity) back to Layer 1.
- Layer 1 stores or verifies that data, providing security and a final source of truth.
This separation lets Layer 2 networks process thousands of transactions per second at a fraction of the cost while still depending on the underlying chain to keep things honest.
Examples of Layer 2 blockchains:
- On Ethereum: Arbitrum, Optimism, Base, zkSync Era, Starknet, Linea
- On Bitcoin: Lightning Network (state channels), Stacks, Rootstock (sidechain-style), Merlin Chain (rollup-style)
Pros | Cons |
| ✅ Lower transaction fees | ✖ Adds another layer of complexity for users |
| ✅ Higher throughput than the base chain | ✖ Most are still partially centralized (e.g. single sequencer) |
| ✅ Inherits security from Layer 1 (to varying degrees) | ✖ Some require waiting periods to withdraw back to L1 |
| ✅ Faster confirmation for end users | ✖ Bridge risks if assets need to move between chains |
Key Differences Between Layer 1 and Layer 2
| In short: The split between Layer 1 and Layer 2 is not just about speed and fees – it cuts across six distinct dimensions that matter for how transactions are processed, secured, and stored. |
Scalability & throughput
Layer 1 throughput is bounded by the network's own consensus design.
- Bitcoin processes roughly 7 transactions per second
- Ethereum mainnet handles about 15
- Solana sits at the other end of the L1 spectrum, with a theoretical capacity of 65,000 TPS thanks to parallel execution.
Layer 2s, by offloading execution from the base chain, can comfortably push past 1,000 TPS – and ZK rollups, in theory, can scale to tens of thousands.
The important caveat: the headline TPS figures most chains advertise are theoretical peaks. Real-world performance under congestion is usually a fraction of the marketed number.
Transaction costs (gas fees)
Layer 1 fees scale with demand for block space.
- On Ethereum, gas fees historically reached $50 to $100 per transaction during peak activity.
- On Bitcoin, they spike when block space is contested.
Layer 2 fees are much lower because the cost is split between the L2's own execution and a small share of L1 data posting. After Ethereum's Dencun upgrade in March 2024, Layer 2 fees dropped sharply.
Starknet's median transaction cost fell from about $1.35 to $0.0196, a 98% decrease, and the broader rollup ecosystem saw similar reductions. The trade-off: L2 fees can still spike when blob demand outpaces supply, since rollups now compete for a separate but limited fee market.
Security model
Layer 1 security is self-contained. The network's validators (or miners) defend it directly, and there is no outside party to fall back on.
Layer 2 security is inherited, but how completely depends on the design.
- A rollup that posts both transaction data and validity proofs to Layer 1 inherits something close to the full security of the base chain.
- A sidechain with its own validator set inherits much less.
The L2Beat framework formalizes this with three stages:
- Stage 0 ("Full Training Wheels," largely operated by the team)
- Stage 1 ("Limited Training Wheels," with a Security Council and functional proofs)
- Stage 2 ("No Training Wheels," fully governed by smart contracts).
Most production L2s are still at Stage 0 or 1 as of 2026.
Decentralization trade-offs
Layer 1 chains aim for maximum decentralization at the consensus layer: thousands of validators, geographic distribution, and open participation.
Layer 2s often trade decentralization for performance. Many major rollups still operate with a single sequencer – a centralized entity that orders transactions before posting them to Layer 1.
That single sequencer is a point of liveness risk: if it goes down or censors transactions, users can be temporarily blocked from transacting until they exit through the slower, on-chain forced inclusion path.
Execution vs settlement responsibilities
This dimension is the cleanest way to understand the relationship between layers.
- Layer 1's job is settlement: storing the canonical record, verifying proofs, and providing finality.
- Layer 2's job is execution: actually running transactions, applying state changes, and managing the user experience.
Splitting these responsibilities is what makes the modern blockchain stack possible. Layer 1 doesn't need to be fast – it just needs to be secure and reliable. Layer 2 doesn't need to be ultra-decentralized – it just needs to anchor its work in a chain that is.
Data availability and state growth
Every transaction creates state, and storing that state on Layer 1 forever is expensive.
Before Ethereum's Dencun upgrade, rollups had to post their data to Ethereum as calldata, which was permanent and costly – adding up to roughly $34 million per month across the ecosystem.
Dencun introduced blob transactions, which let rollups post temporary data that is verified once and discarded after about 18 days. This massively reduced costs for L2s but also raised a new question: where does the data live long-term?
Some rollups use external data availability layers (Celestia, EigenDA), which makes them cheaper but introduces new trust assumptions.
When Each Layer Is Put Into Practice
| In short: The clearest way to see how Layer 1 and Layer 2 fit together is to follow a transaction through the stack. Both layers are usually involved, even when the user only sees one. |
Consider a typical token swap on Arbitrum, an Ethereum Layer 2:
- The user signs the transaction in their wallet and sends it to the Arbitrum network.
- Arbitrum's sequencer orders the transaction and executes it almost instantly, deducting tokens and updating balances on the L2.
- The user sees their swap confirmed in a few seconds.
- Behind the scenes, Arbitrum batches this transaction together with thousands of others.
- That batch is compressed and posted to Ethereum L1 as a blob, becoming part of Ethereum's verified history.
- Once posted (and after a challenge window for optimistic rollups), the transaction is considered fully settled.
The user experiences a fast, cheap swap. Ethereum L1 records the underlying data and provides the security guarantee that the swap actually happened as described.
>> Learn more: What Is Ethereum Finality: When Transactions Are Truly Final
When users and builders interact with each layer
In day-to-day usage, the dividing line is roughly:
- Layer 1 is where assets are anchored: deposits and withdrawals from centralized exchanges, large transfers, and high-value settlement typically happen on the base chain.
- Layer 2 is where activity happens: DeFi swaps, NFT minting, gaming transactions, payments, and most consumer-facing dApps live on L2s because the fees and speed make them viable.
For builders, the choice depends on the use case.
- Applications that need high throughput, low cost, or frequent interactions (a perpetuals DEX, an on-chain game, a payment app) almost always pick a Layer 2.
- Applications that prioritize maximum security, simplicity, or institutional alignment (a Bitcoin treasury product, a high-value bridge, a settlement protocol) tend to stay on Layer 1.
How Layer 2 Differs Across Blockchains
| In short: One of the most common sources of confusion is treating "Layer 2" as a single, well-defined category. In practice, what counts as a Layer 2 looks very different depending on which Layer 1 you are looking at. |
- On Ethereum: "Layer 2" has come to mean rollups – networks that execute transactions off-chain and post compressed data and proofs back to Ethereum for verification. The L2Beat framework defines clear technical requirements: an L2 must publish all transaction data to Ethereum and allow users to reconstruct state independently.
- On Bitcoin: "Layer 2" historically meant payment channels like the Lightning Network. Lightning lets two parties open a channel, transact off-chain at near-zero cost, and only settle the final state on Bitcoin.
- On Solana: The Layer 2 conversation barely exists. Solana scales at Layer 1 through parallel execution via its Sealevel runtime, processing thousands of transactions per second on the base chain itself. There's no architectural pressure to push activity to a separate layer, and few teams have tried.
→ The takeaway is that "Layer 2" describes the role a network plays in a larger stack – handling execution while another chain handles settlement – rather than a specific technology. Two networks can both be called Layer 2 and have almost nothing in common under the hood.
The Blockchain Trilemma in 2026 – Has It Been Solved?
| In short: The blockchain trilemma – the idea that any blockchain must trade off between scalability, security, and decentralization – has not been solved. But the way the industry handles it has fundamentally changed. |
The trilemma was first popularized by Vitalik Buterin around 2017, arguing that a blockchain can comfortably achieve at most two of the three properties: be highly scalable, highly secure, or highly decentralized.
For most of crypto's history, this was treated as a hard architectural ceiling that any single chain had to negotiate.
The Layer 1/Layer 2 split changed the framing. Instead of trying to solve the trilemma on one network, the modern stack outsources it:
- Layer 1 holds security and decentralization.
- Layer 2 absorbs the scalability demand.
Each layer wins two of the three properties at its level, and the overall system covers all three.
This sounds elegant, but there's a catch. If a Layer 2 is itself not decentralized – running on a single sequencer, with a small multisig controlling upgrades, then the combined stack is not really winning the decentralization side of the trilemma. It just relocated the trade-off.
According to L2Beat's reporting, most production rollups in 2026 are still at Stage 0 or Stage 1, meaning they still rely on Security Councils or training wheels.
→ So the honest answer is: the trilemma has not been solved, but it has been split across layers. Whether that counts as a real solution depends on how seriously each layer treats its share of the responsibility.
The Future of Blockchain Scaling
| In short: The Layer 1/Layer 2 vocabulary is starting to feel cramped. Several developments in 2024–2026 are reshaping how blockchains are built and how scaling responsibility is allocated. |
- Modular blockchains are unbundling the four core jobs of a blockchain – execution, settlement, consensus, and data availability – into separate, specialized layers. Projects like Celestia focus exclusively on providing data availability for other chains to consume.
- Ethereum's roadmap doubles down in this direction. The Dencun upgrade in March 2024 introduced blob transactions, which gave rollups a dedicated, cheaper data lane. The follow-up upgrades – Pectra in May 2025 and Fusaka in December 2025 – pushed blob capacity higher and improved the long-term economics for both Ethereum and L2s.
- App-chains and sovereign rollups are growing in popularity, especially for use cases that need full control over their execution environment. Projects like dYdX v4 chose to launch as a Cosmos app-chain rather than as an Ethereum L2, citing the need for ordered transaction sequencing and chain-level control over MEV.
- Restaking and shared security are blurring the line between layers. Through systems like EigenLayer, ETH stakers can extend Ethereum's security to other services – not just rollups, but also data availability layers, oracles, and bridges.
The likely outcome is that "Layer 1" and "Layer 2" will eventually feel more like positions in a stack that has more than two slots. A future blockchain might be best described by the function it performs: an execution chain, a data availability layer, a settlement chain.
Sources and Further Reading
- L2Beat – The State of the Layer 2 Ecosystem
- L2Beat – Stages Framework for Rollup Maturity
- Ethereum.org – Layer 2 Networks
- Vitalik Buterin – Endgame (Rollup-Centric Roadmap)
- EIP-4844 Specification – Shard Blob Transactions
- Wikipedia – Blockchain Scalability
- DL News – Ethereum's Dencun Upgrade Lowers Layer 2 Fees by as Much as 98%
FAQs About Layer 1 vs Layer 2 Blockchain
No. A Layer 2 fundamentally depends on a Layer 1 for security and final settlement. If the underlying L1 goes offline or suffers a critical failure, the L2 cannot finalize transactions or process trustless withdrawals. This is what gives L2s their security — but also why they are not truly independent networks.
