A Layer 1 (L1) blockchain is the foundational settlement layer of a blockchain network. It maintains its own validator or miner set, enforces its own consensus rules, and settles transactions with no external chain providing security. Bitcoin, Ethereum, Solana, and Avalanche are all L1 networks. Every higher-order system built on top -- rollups, sidechains, app-chains -- ultimately roots its security in one or more L1s.
​

The defining characteristic of an L1 is that it does not borrow security from another chain. When a Bitcoin miner appends a block, the proof-of-work embedded in that block is verified by every full node on the network. No outside party vouches for the result. This self-contained trust model is what makes an L1 the "base layer" in any multi-tier architecture. See what trustless means in blockchain for a deeper look at what this independence actually guarantees.
​

L1 blockchains carry three core responsibilities:

As blockchains have matured, some L1s have deliberately offloaded execution to Layer 2 networks while retaining data availability and settlement at the base layer. Ethereum's rollup-centric roadmap is the clearest example of this architectural choice. Ethereum L1 processes roughly 15-30 TPS at the base layer -- a deliberate constraint that keeps block space scarce, pushes fees to rollups, and maintains a low hardware bar for running a full node.
​

Full-node counts give a rough proxy for decentralization: Ethereum has over 7,000 non-validator full nodes; Bitcoin maintains roughly 15,000-17,000 reachable nodes. Both figures are orders of magnitude higher than most permissioned ledgers, which matters because the node set is what makes censorship of any single participant economically impractical.
​

The base layer concept and how Ethereum defines it are documented at ethereum.org's introduction to Ethereum.
​

Vitalik Buterin's blockchain trilemma states that a decentralized blockchain can optimize for at most two of three properties at once: security, scalability, and decentralization. Maximizing all three simultaneously requires architectural trade-offs -- which explains why no single L1 has become the universal settlement layer for all use cases, and why L2s exist at all.
​

The trilemma works like this. Security means the chain can resist attacks even from well-resourced adversaries. Decentralization means a large, geographically distributed node set where no minority can censor or revert transactions. Scalability means the chain can process many transactions cheaply and quickly.
​

Each major L1 lands somewhere different on this triangle:

Understanding where each L1 sits in the trilemma is essential context for why Layer 2 networks exist. L2s effectively move the scalability leg offchain while inheriting the security of the L1 settlement layer underneath them. The trilemma framing itself is discussed in Buterin's 2021 sharding writeup.
​

Consensus is the mechanism by which independent nodes in a blockchain network agree on a single canonical history of transactions without trusting each other. Different L1s use fundamentally different approaches -- proof-of-work, proof-of-stake, proof-of-history, and probabilistic sampling -- each with distinct security assumptions, energy profiles, and finality characteristics.
​

Bitcoin's proof-of-work requires miners to find a hash below a target difficulty by brute-force iteration. The computational work burned in this process is what makes rewriting history expensive: to reorg a chain of N blocks, an attacker must redo all that work faster than the honest chain grows. At Bitcoin's current hashrate, reversing 6 confirmed blocks would require more electricity than most countries consume in a day. The longest-chain rule (Nakamoto consensus) means nodes always follow the chain representing the most accumulated work. The security model is documented in the original Bitcoin whitepaper.
​

Ethereum moved to proof-of-stake with The Merge in September 2022. Validators lock 32 ETH as collateral. The protocol selects proposers and attesters pseudorandomly, weighted by stake. Validators who sign conflicting blocks or go offline are slashed -- losing a portion of their staked ETH. Casper FFG, the finality gadget layered on top of LMD GHOST fork-choice, provides deterministic finality: once a checkpoint is finalized (two epochs, roughly 12-15 minutes), it cannot be reverted without destroying at least one-third of the total staked ETH. The full consensus mechanism is documented at ethereum.org's consensus mechanisms page.
​

Solana uses proof-of-history (PoH) as a verifiable delay function that creates a cryptographic timestamp for every event before it enters consensus. This lets validators agree on event ordering without a separate round of message-passing. PoH is not itself a consensus mechanism -- it runs alongside Tower BFT, Solana's version of PBFT, to achieve fast block times (400ms slots). The combination is what enables Solana's high theoretical throughput, though real-world sustained TPS has typically been lower, around 2,000-4,000 TPS under normal load. Solana's architecture is documented at solana.com's proof-of-history docs.
​

Avalanche uses a family of consensus protocols (Snowflake, Snowball, Avalanche) based on repeated random subsampling. A node queries a small random sample of other validators about the preferred outcome. If a supermajority agrees, the node adopts that preference. Iterating this process rapidly converges on a single answer with probabilistic guarantees. Avalanche consensus achieves finality in under two seconds on the primary network and scales to thousands of validators without the communication overhead of classical BFT, because each round involves only a random subset. The mechanism is explained at Avalanche's documentation.
​

A Layer 1 blockchain settles transactions using its own consensus and node set. A Layer 2 is a separate execution environment that processes transactions offchain, then posts compressed state proofs or transaction data back to the L1 for final settlement. The L2 does not run its own independent validator set for security -- it inherits the L1's security guarantees by anchoring its state onchain.
​

The practical consequence is that L2s can offer far higher throughput and lower fees than their underlying L1, without introducing new trust assumptions at the settlement layer. An Optimism or Arbitrum transaction ultimately settles on Ethereum L1 -- if the L2 operator disappears tomorrow, users can always exit by submitting proof of their balance directly to the Ethereum mainnet contract.
​

Key structural differences:
​

Rollups are the dominant L2 design today. Optimistic rollups post transaction data to L1 and assume validity unless a fraud proof is submitted within a challenge window. ZK rollups post a cryptographic validity proof that mathematically guarantees correctness, enabling faster withdrawals. For a thorough comparison, see What Is a Rollup? Optimistic vs ZK Rollups.
​

Beyond L2, there is now a growing layer of Layer 3 networks that build on top of L2s, adding application-specific customization without the cost of settling directly on L1. L2Beat tracks the security status and total value locked across all major L2 deployments at l2beat.com.
​

Finality describes the point at which a transaction can no longer be reversed. L1s differ sharply on how they define it: Bitcoin uses probabilistic finality, where safety grows with each additional block; Ethereum's Casper FFG and Avalanche's consensus both provide deterministic (or near-deterministic) finality within a bounded time window. This distinction matters most for settlement-sensitive applications like stablecoin payments and cross-chain bridges.
​

Bitcoin does not have a built-in finality mechanism. Every new block added on top of a transaction makes it exponentially harder to reorg, but the possibility is never formally zero at the protocol level. The 6-block confirmation standard (roughly 60 minutes at a 10-minute average block time) is an industry convention, not a protocol rule. The underlying assumption is that an attacker would need to control more than 50% of the global hashrate and sustain that advantage for six blocks. Satoshi Nakamoto's whitepaper derives the probability tables: after 6 blocks, the chance of a successful double-spend by an attacker with 10% hashrate is below 0.1%.
​

Ethereum's Casper FFG checkpoint mechanism finalizes epochs (32-slot periods, roughly 6.4 minutes each) once two consecutive checkpoints each collect attestations from more than two-thirds of the total stake. A finalized checkpoint cannot be reversed without slashing at least one-third of all staked ETH -- over 10 million ETH at current levels. This is deterministic in the sense that violating it requires a protocol-level, economically catastrophic act, not just raw hashrate. Most Ethereum applications treat a transaction as "soft finalized" after one block (roughly 12 seconds) for user experience purposes; for cross-chain bridges and large settlements, waiting for Casper finality provides the strongest guarantee.
​

Avalanche's sub-two-second finality stems from its probabilistic subsampling consensus reaching a confidence threshold. Once the protocol's confidence counter crosses the threshold, the transaction is considered final. In practice the Avalanche C-Chain advertises under-two-second finality for normal network conditions, which is among the fastest of any permissionless L1.
​

The significance of finality for applications built on top of L1s is explored further in What Does Trustless Mean in Blockchain? Ethereum's finality model in detail is at ethereum.org's Gasper documentation.
​

Five networks dominate L1 activity by total value locked, developer activity, and stablecoin issuance as of early 2026: Ethereum, Bitcoin, Solana, Avalanche, and BNB Chain. Each occupies a distinct niche shaped by its consensus model, fee market, and ecosystem of applications built on top of it.
​

Ethereum is the dominant smart-contract L1 by TVL and developer count. Its EVM (Ethereum Virtual Machine) became the de facto standard for smart contracts, which is why nearly every new L1 and L2 ships EVM compatibility. Post-Merge, Ethereum runs on proof-of-stake with over 1 million active validators as of 2025. Native throughput at L1 sits around 15-30 TPS. The deliberate design choice is to keep L1 block space scarce and expensive, pushing high-volume activity to rollups. ETH is the native token used for gas fees and validator staking. See ethereum.org for current protocol documentation.
​

Bitcoin is the oldest and most secure blockchain network, optimized for a single use case: censorship-resistant transfer of value. Its scripting language is intentionally limited. Bitcoin processes roughly 7 TPS onchain. The Lightning Network is a payment-channel L2 that routes micropayments offchain, but Bitcoin's L1 does not support general-purpose smart contracts. The 21-million BTC supply cap and the halving cycle (most recent halving was April 2024, reducing block reward to 3.125 BTC) are hard-coded into the protocol. Bitcoin L1 holds the largest share of total crypto market capitalization. Protocol specifications are at bitcoin.org's developer guide.
​

Solana is designed for high-throughput consumer and payments applications. Its combination of proof-of-history and Tower BFT targets 400ms slot times. The Solana Foundation claims theoretical peak throughput of 65,000 TPS; real-world sustained throughput under normal load has been in the 2,000-4,000 TPS range. Solana suffered multiple network halts between 2021 and 2022 during congestion events, which prompted a series of network upgrades including QUIC networking and stake-weighted QoS. The validator hardware requirements are higher than Ethereum's, which has fueled ongoing debate about validator set concentration. SOL is the native token. See solana.com for current network statistics.
​

Avalanche's primary network runs three chains: the X-Chain (asset transfers using Avalanche consensus), P-Chain (validator and subnet coordination), and C-Chain (EVM-compatible smart contracts). Most DeFi activity runs on the C-Chain. Avalanche's subnet architecture allows projects to launch application-specific chains that share the validator set of the primary network or define their own. AVAX is the native token, used for gas on the C-Chain, subnet creation deposits, and validator staking. Finality is under two seconds on the C-Chain. Avalanche documentation is at docs.avax.network.
​

BNB Chain (formerly Binance Smart Chain) is an EVM-compatible L1 operated by Binance's ecosystem. It uses a delegated proof-of-stake model called Proof of Staked Authority (PoSA) with a validator set historically capped at 21 active validators, recently expanded to 40. This smaller validator set enables fast block times (around 3 seconds) and low fees, but with a more concentrated trust model than Ethereum or Bitcoin. BNB is the native token. BNB Chain is the highest-volume EVM chain by transaction count during periods of high retail activity, largely because of its low fees. Network data is available at BscScan.
​

The five major L1 networks differ across consensus model, native throughput, finality speed, EVM compatibility, and token economics. The table below consolidates the key parameters for direct comparison. TPS figures are sourced from official documentation and developer benchmarks; sustained real-world throughput is typically lower than theoretical peaks, and varies with network load.
​

A few notes on reading this table. Solana's TPS range is wide because the claimed 65,000 figure is a theoretical maximum under ideal conditions; sustained real-world throughput varies significantly with network congestion and application mix. Ethereum's L1 TPS appears low precisely because of the rollup-centric design -- the relevant aggregate throughput figure for the Ethereum ecosystem is the sum across all its L2s (Arbitrum, Optimism, Base, etc.), which collectively process hundreds of TPS. For current L2 data, see the rollup comparison article or the live tracker at L2Beat.
​

Circle issues USDC natively (not as a bridged wrapper) on a growing list of L1s via its Cross-Chain Transfer Protocol (CCTP). Native issuance means Circle mints and burns USDC directly on the chain using a canonical smart contract -- eliminating the bridge-wrapper risk that comes with cross-chain lock-and-mint designs. As of early 2026, the L1 networks with native USDC issuance include Ethereum, Solana, and Avalanche, along with several L2s including Base.
​

Why this matters: applications that require institutional-grade stablecoin settlement prefer native USDC because there is no third-party bridge contract between the user and Circle's redemption guarantee. For DeFi protocols handling large flows, bridge-wrapped USDC (sometimes labelled USDC.e or wUSDC) introduces smart-contract risk in the bridge layer. Native USDC removes that risk.
​

CCTP works by allowing Circle to burn USDC on the source chain and attest the burn onchain, then mint an equivalent amount on the destination chain. The attestation lives onchain on both networks, meaning the cross-chain transfer is verifiable without trusting a multisig bridge. CCTP is available on Ethereum, Avalanche C-Chain, Solana, and several L2s including Base, Arbitrum, Optimism, and Polygon.
​

For stablecoin automation workflows that span multiple L1s and L2s, the best stablecoin automation platforms guide covers the current landscape of tools built for multi-chain treasury and payment flows. USDC-supported chains and CCTP documentation are maintained at circle.com/en/cross-chain-transfer-protocol.
​

Ethereum is an L1 blockchain. It maintains its own validator set (over 1 million active validators post-Merge) and settles transactions using its own proof-of-stake consensus. Rollups like Arbitrum, Optimism, and Base are L2s that settle on Ethereum, but Ethereum mainnet itself is the base settlement layer with no external security dependency.
​

Decentralization is a function of how many independent nodes participate in consensus, the minimum hardware required to run one, and the geographic and political distribution of validators. Chains with high hardware requirements (Solana) or small validator sets (BNB Chain at 21-40 validators) score lower on decentralization than Ethereum or Bitcoin, where running a full node is feasible on consumer hardware with modest storage.
​

Yes. Every L1 has a maximum throughput defined by its block size and block time. When demand exceeds capacity, a fee market emerges -- users bid for scarce block space. Ethereum's EIP-1559 introduced a base fee that adjusts dynamically with demand. On Bitcoin, periods of high inscription activity have pushed mempool fees significantly higher, creating backlogs for lower-fee transactions waiting for inclusion in a block.
​

L2s depend on L1s for their security guarantees. If Ethereum L1 were compromised or shut down, every rollup settling on it would also be compromised. The L1 is the root of trust -- it is where fraud proofs are adjudicated, where validity proofs are verified, and where users can always exit even if the L2 operator goes dark. L2s are optimizations on top of L1 security, not replacements for it. See What Is a Layer 2 (L2) Blockchain? for the full picture.
​

Avalanche subnets are L1 deployments in their own right: they run their own validator set (which may overlap with Avalanche's primary validators) and use Avalanche consensus internally. An L2 like an Ethereum rollup posts state to its parent L1 and inherits security from it. Subnets are sovereign chains; L2s are not. The distinction matters for security assumptions: a subnet does not inherit Ethereum's security, while an L2 rollup does.
​

Related reading