Layer 2 Scaling Solutions Complete Guide 2026

Introduction

Ethereum’s success has been paradoxically limited by its own popularity. As adoption grew, network congestion drove transaction costs to prohibitive levels. Layer 2 scaling solutions emerged as the critical pathway to solving these constraints while maintaining Ethereum’s security and decentralization guarantees.

By 2026, Layer 2 has transformed from experimental infrastructure into the dominant way users interact with Ethereum. Combined TVL across L2s surpasses $34 billion, with daily transactions consistently dwarfing Ethereum mainnet. The ecosystem has undergone a dramatic consolidation — Base, Arbitrum, and Optimism now process nearly 90% of all L2 transactions, while dozens of smaller rollups have faded into zombie chain status.

The Dencun upgrade (EIP-4844) in 2024 was a watershed moment, slashing L2 data costs by roughly 10x and making transactions under $0.01 standard. This cost reduction unlocked new use cases — social applications, gaming, microtransactions — that were economically unviable before. The competition has since shifted from who offers the cheapest fees to who provides the best developer stack, deepest liquidity, and most compelling user experience.

This guide provides comprehensive coverage of the Layer 2 landscape as it stands in mid-2026. We’ll examine the fundamental scaling approaches, explore the leading implementations — Arbitrum, Base, Optimism, zkSync Era, and Starknet — analyze their architectural tradeoffs, and consider the future evolution of this critical infrastructure.

Understanding Layer 2 Architecture

The Rollup-Centric Roadmap

Ethereum’s scaling strategy centers on rollups — Layer 2 constructions that execute transactions off-chain while posting compressed data to Ethereum mainnet. This approach preserves Ethereum’s security guarantees while dramatically improving throughput and reducing costs.

The roadmap has largely been validated. By 2026, most user activity occurs on Layer 2 networks, with Ethereum mainnet serving primarily as a settlement and data availability layer. This “hyperchain” vision — Ethereum as the settlement backbone for a vast ecosystem of specialized L2s — is increasingly the operating reality.

Optimistic vs. Zero-Knowledge Rollups

The two primary rollup approaches differ fundamentally in their verification mechanisms. In 2026, optimistic rollups command roughly 90% of L2 TVL, while ZK rollups hold approximately 10%.

Optimistic rollups assume transactions are valid by default, posting data without on-chain proof. Invalid transactions can be challenged during a dispute period — typically seven days — during which anyone can submit a fraud proof. Arbitrum’s BoLD (Bounded Liquidity Delay) protocol, live since February 2025, improved this mechanism by enabling permissionless validation with bounded dispute times. This approach is simpler to implement but requires the honest minority assumption.

ZK rollups generate cryptographic proofs — validity proofs — that mathematically prove the correctness of state transitions. Each batch includes a proof verifiable on-chain, providing immediate finality without a challenge period. ZK rollups offer stronger security guarantees but remain more technically complex and computationally intensive. The proving gap has narrowed significantly, however, with advances like StarkWare’s Stwo prover reducing proof sizes by 94% (from 1.3MB to 77KB).

The Dencun upgrade (EIP-4844) in 2024 transformed the economics for both approaches by introducing blob data, reducing L2 posting costs by approximately 10x. Transaction fees on most L2s now routinely sit below $0.01 for simple transfers and $0.10-$0.30 for token swaps.

Leading Layer 2 Protocols

Arbitrum: The DeFi Heavyweight

Arbitrum remains the dominant Layer 2 by total value secured, with approximately $16.7 billion TVS and a market share of roughly 40% of L2 DeFi TVL. It has benefited from early mover advantage, deepest DeFi liquidity, and strong EVM equivalence.

Multiple major upgrades have kept Arbitrum competitive. The BoLD dispute protocol (February 2025) enabled permissionless validation with bounded dispute times, replacing the previous whitelisted-validator model. ArbOS 40 “Callisto” (June 2025) introduced native account abstraction. ArbOS 51 “Dia” (January 2026) revamped the fee market with multi-window EIP-1559-style pricing, improving gas fee stability during demand spikes — a dynamic pricing system that uses six target-window pairs to smooth out fee volatility rather than allowing a single target to behave like a springboard.

Stylus, Arbitrum’s mechanism for writing smart contracts in Rust, C, and C++, has expanded the developer ecosystem beyond Solidity. This enables developers to bring existing codebases and performance-critical applications to Arbitrum without learning new languages.

The Orbit framework allows anyone to launch custom L2 or L3 chains that settle to Arbitrum, with configurable parameters including gas token, TPS limits, and governance. Notable Orbit chains include Xai (gaming) and a planned Robinhood blockchain for tokenized equity trading.

Arbitrum employs an optimistic rollup with the AnyTrust guarantee — security as long as at least one honest validator exists. Arbitrum Nova, a separate chain using a data availability committee, targets gaming and social applications requiring ultra-low costs. Timeboost, a new transaction ordering mechanism, gives users the ability to pay for priority inclusion while capturing MEV for the protocol rather than leaving it to searchers.

Base: The Consumer Powerhouse

Base, launched by Coinbase in August 2023, is the breakout story of the L2 landscape. By 2026 it leads all L2s in daily active users and transaction count, processing over 11 million transactions daily with 663,000+ active addresses. Base’s TVL has reached approximately $8-11 billion, and it commands over 46% of L2 DeFi TVL — making it the largest L2 by this metric.

Base’s success is driven by Coinbase’s massive distribution advantage — over 110 million verified users with seamless onboarding. For many crypto newcomers, Base is their first L2 experience. The Base App, an all-in-one platform combining wallet, trading, payments, social networking, and DeFi, is positioning itself as an on-chain super app.

The network’s Flashblocks technology delivers 200ms block times, making the user experience competitive with centralized platforms. Base achieved Stage 1 decentralization status and has become a hub for consumer applications, social finance, real-world asset tokenization, and institutional on-chain activity — capturing 43% of weekly on-chain BTC spot volume.

Base is built on the OP Stack, though Coinbase announced in February 2026 that it would begin migrating off the standard stack toward greater sovereignty. A native Base token is being explored, with Polymarket odds suggesting a 69% probability of launch in 2026.

Optimism: The Superchain Architect

Optimism has evolved from a single L2 chain into the architect of the Superchain — a network of OP Stack-based chains sharing security, bridges, and governance. As of mid-2026, the Superchain includes OP Mainnet, World Chain (Tools for Humanity), Mode, Zora, Soneium, Ink (Kraken), Unichain (Uniswap), and over a dozen other production chains. The OP Stack now powers over 62% of all L2 transactions.

OP Mainnet itself holds approximately $5.6 billion TVL. However, the Superchain’s collective assets are substantially larger when including chains like Base and World Chain. The OP Stack’s modular, open-source architecture makes launching Ethereum-aligned L2 chains accessible to any team, and has become the default framework for enterprise and institutional L2 deployments.

In January 2026, Optimism governance approved a landmark OP token buyback program: 50% of net Superchain sequencer revenue will be directed toward recurring OP buybacks over a 12-month pilot. This directly links the OP token’s value to the economic performance of the Superchain ecosystem.

Native interop is the next major milestone. A testnet for cross-chain message passing and atomic composability between Superchain chains is already live. This would enable seamless asset transfers, unified liquidity, and cross-chain contract calls within the Superchain — addressing the fragmentation that has long plagued multi-chain architectures.

Optimism’s retroactive public goods funding (RetroPGF) mechanism remains unique in the space, directing protocol revenue to projects that provide public goods to the ecosystem based on community voting.

zkSync Era: The Elastic Network

zkSync Era, developed by Matter Labs, has shifted focus from a single ZK rollup to the Elastic Network — a network of ZK chains that share liquidity and interoperability through ZKsync Gateway. The original ZKsync Lite was officially deprecated on May 4, 2026, after serving as a proof-of-concept that validated critical zero-knowledge rollup designs.

The zkEVM implementation achieves EVM compatibility while generating ZK proofs — a significant technical achievement, though as a Type 4 zkEVM, some Ethereum bytecode quirks require adjustment. The ZK Stack (powered by the Atlas upgrade) enables anyone to launch custom ZK chains with shared proving infrastructure.

The ZKsync Gateway upgrade has expanded interoperability capabilities beyond simple message passing to include asset transfer support and interop bundles. In early 2026, ZKsync Era temporarily migrated to settling directly on Ethereum L1 during the Gateway upgrade rollout.

zkSync Era holds approximately $404 million in TVS. While significantly smaller than the leading optimistic rollups, the ZK value proposition — immediate finality, stronger cryptographic guarantees — continues attracting specific use cases where these properties matter most.

Starknet: STARKs for Scale

Starknet employs STARKs (Scalable Transparent Arguments of Knowledge), offering advantages in transparency (no trusted setup) and quantum resistance. The 2025-2026 “Quantum Leap” upgrades have brought sub-second block times through a high-speed sequencer, making trading experiences on Starknet DEXs competitive with centralized platforms.

STRK staking went live in 2026, allowing holders to delegate tokens to sequencers and provers, earning rewards while securing the network. The Stwo prover achieved a dramatic milestone — reducing Cairo proofs from 1.3MB to 77KB, removing the proof size bottleneck that previously limited STARK verification on other chains.

Kakarot, a ZK-EVM running on Starknet, enables Ethereum developers to deploy existing Solidity dApps on Starknet without modification, bridging the gap between Cairo-native and EVM ecosystems. The strkBTC launch in May 2026 brought private Bitcoin to Starknet DeFi, supporting shielded balances and productive BTC use through the STRK20 standard.

Starknet holds approximately $578 million TVS. Bitwise filed for a STRK ETF in December 2025, which would become the first STRK ETF if approved, signaling growing institutional interest in ZK-native infrastructure.

Technical Deep Dive

Sequencer Architecture

The sequencer receives user transactions, orders them, and submits batches to mainnet. Sequencer design directly determines a Layer 2’s security properties and decentralization.

Most L2s still employ a single permissioned sequencer run by a single operator. This centralized design improves performance — Base’s Flashblocks achieve 200ms block times — but creates trust assumptions around censorship resistance and liveness. The shared-sequencer landscape has faced setbacks: Astria shut down entirely in 2025, highlighting how early-stage this category remains.

The decentralization roadmap for most protocols includes transitioning to permissionless validator sets. Arbitrum’s roadmap targets a decentralized fair sequencing model where a committee collectively determines ordering. The Superchain’s native interop roadmap includes shared sequencing, which would enable atomic cross-chain transactions within the Superchain. However, achieving decentralized sequencing at scale remains one of the hardest open problems in L2 engineering.

Data Availability

Data availability determines whether users can reconstruct the Layer 2 state and withdraw funds if the operator disappears. For a deeper dive into data availability layers and the modular blockchain thesis, see our complete guide to data availability layers.

On-chain data availability posts all transaction data to Ethereum, maximizing security. Dencun’s blob data (EIP-4844) dramatically reduced the cost of this approach by introducing a separate fee market for data blobs. Ethereum’s upcoming Glamsterdam upgrade (planned 2026) will further expand blob throughput with enshrined proposer-builder separation (ePBS), setting the stage for another step-change in L2 capacity — potentially supporting 100,000+ TPS across the entire L2 ecosystem.

Off-chain data availability stores data elsewhere, with the L1 contract receiving only commitments. This reduces costs but introduces trust assumptions. Arbitrum Nova and zkSync Era’s legacy zkPorter feature use off-chain DA for applications where lower security assumptions are acceptable in exchange for minimal fees.

Validium uses ZK proofs for correctness with off-chain data availability, providing an intermediate security-cost tradeoff.

Bridge Architecture

Bridges enable asset movement between Layer 1 and Layer 2, and across different L2s. Bridge security remains critical — bridge exploits have cost billions. For a comprehensive overview of bridge designs and security models, see our guide to cross-chain interoperability and bridges.

The canonical bridge — managed by the L2 protocol — provides the most direct path for cross-layer movement. Users deposit to L1 contracts, and L2 contracts mint corresponding assets. Withdrawals in optimistic rollups require waiting through the challenge period (typically seven days, though users can pay for faster exits via liquidity providers). ZK rollups provide faster withdrawals since validity proofs can be verified immediately on L1.

Third-party bridges remain popular for speed and convenience but introduce additional trust assumptions. The trend in 2026 is toward canonical interoperability — native cross-chain message passing without third-party bridge intermediaries, as exemplified by Optimism’s Superchain interop and ZKsync’s Elastic Network.

Ecosystem Analysis

DeFi on Layer 2

DeFi has migrated en masse to Layer 2, driven by the cost differential — a Uniswap swap that costs $5-50 on Ethereum mainnet costs $0.10-0.30 on Arbitrum or Base. Major protocols including Uniswap, Aave, Morpho, and Pendle maintain deployments across multiple L2s, enabling users to choose their preferred network while accessing similar financial primitives.

Cross-chain DeFi has emerged as a significant use case. Applications leverage liquidity across multiple L2s through canonical bridges and specialized cross-chain DEX protocols. Aerodrome on Base, Velodrome on Optimism, and GMX on Arbitrum exemplify the L2-native DeFi protocols that have captured substantial market share from their L1 predecessors.

Real-world asset (RWA) tokenization has particularly flourished on L2s. The combination of low fees, Ethereum-grade security, and institutional tooling has made networks like Arbitrum and Base preferred venues for tokenized Treasuries, private credit, and commodities. BlackRock’s BUIDL fund and similar products are accessible through L2 DeFi composability.

Consumer Applications and SocialFi

Consumer applications have emerged as the fastest-growing L2 segment, led by Base. The combination of sub-cent fees, fast confirmation times, and Coinbase’s distribution has made Base a hub for social applications, prediction markets, and creator economies. Zora (operating as an OP Stack chain) has pioneered on-chain media with cheap NFT minting. Polymarket and similar prediction markets see significant volume on L2s.

Gaming

Blockchain gaming has particularly benefited from L2 scaling. The high transaction frequency of gaming — every action potentially requiring a blockchain transaction — makes L1 costs prohibitive. L2s make these use cases economically viable.

Full on-chain games like Sky Strife run entirely on L2 infrastructure, leveraging cost-effectiveness for frequent interactions. Other games use hybrid approaches, maintaining most state off-chain and using L2 only for high-value asset transfers. Arbitrum Nova and specific Orbit gaming chains target this segment with ultra-low-cost infrastructure.

Rollup Types: Optimistic vs Zero-Knowledge

Optimistic Rollups

Optimistic rollups assume transactions are valid unless challenged. They post transaction data to L1 but don’t generate proofs of correctness. A challenge period (typically 7 days) allows verifiers to submit fraud proofs if they detect invalid state transitions.

Pros: EVM-compatible by default, simpler implementation, lower on-chain computation costs Cons: Delayed withdrawals (challenge period), requires honest verifier assumption

// Simplified optimistic rollup fraud proof
contract OptimisticRollup {
    struct StateBatch {
        bytes32 stateRoot;
        uint256 startBlock;
        uint256 endBlock;
        uint256 timestamp;
        uint256[2] proof;
    }
    
    mapping(uint256 => StateBatch) public stateBatches;
    mapping(bytes32 => bool) public challenged;
    
    function submitBatch(bytes32 stateRoot, uint256 startBlock, uint256 endBlock) external {
        uint256 batchId = uint256(keccak256(abi.encodePacked(stateRoot, startBlock, endBlock)));
        stateBatches[batchId] = StateBatch(stateRoot, startBlock, endBlock, block.timestamp, [uint256(0), uint256(0)]);
    }
    
    function challengeBatch(uint256 batchId, bytes32 fraudProof) external {
        require(stateBatches[batchId].timestamp > 0, "Batch not found");
        require(block.timestamp < stateBatches[batchId].timestamp + 7 days, "Challenge period ended");
        // Verify fraud proof and revert batch if valid
        challenged[batchId] = true;
    }
}

Zero-Knowledge Rollups

ZK rollups generate validity proofs that cryptographically verify every state transition. This provides immediate finality without a challenge period.

Pros: Instant withdrawals, stronger security guarantees, no challenge period Cons: More complex proving infrastructure, higher computational overhead for proof generation, EVM compatibility challenges

class ZKRollup:
    def __init__(self):
        self.state = {}
        self.batch_prover = ProvingSystem()
    
    def create_batch(self, transactions):
        # Accumulate state changes
        old_state_root = self.get_state_root()
        for tx in transactions:
            self.apply_transaction(tx)
        new_state_root = self.get_state_root()
        
        # Generate validity proof
        proof = self.batch_prover.generate_proof(
            old_state_root,
            transactions,
            new_state_root
        )
        
        return Batch(new_state_root, proof, transactions)
    
    def verify_batch(self, batch):
        # Verify the ZK proof on-chain
        return self.verifier.verify(
            batch.old_state_root,
            batch.new_state_root,
            batch.proof
        )

Major L2 Economics and Token Models

L2 Interoperability and Cross-L2 Messaging

Current State

The L2 ecosystem remains fragmented. Moving assets between Arbitrum and Optimism typically requires bridging back to Ethereum L1 first, resulting in delays and costs. Several solutions are emerging:

Superchain Interop (Optimism): Native cross-chain message passing between OP Stack chains enables atomic composability across the Superchain ecosystem.

Arbitrum Orbit: Custom L3 chains that settle to Arbitrum can communicate through the AnyTrust protocol.

LayerZero: A universal cross-chain messaging layer that connects L2s with L1 and other chains.

// Cross-L2 messaging via LayerZero
contract CrossL2Messenger {
    function sendMessage(
        uint32 dstChainId,
        address recipient,
        bytes calldata message
    ) external payable {
        // Encode message for destination chain
        bytes memory payload = abi.encode(msg.sender, recipient, message);
        
        // Send through LayerZero endpoint
        ILayerZeroEndpoint(lzEndpoint).send{value: msg.value}(
            dstChainId,
            payload,
            payable(msg.sender),
            address(0x0),
            bytes(""),
            bytes("")
        );
    }
}

Future: Shared Sequencing

Shared sequencers coordinate transaction ordering across multiple L2s, enabling atomic cross-chain transactions. Projects like Espresso and Astria (before its shutdown) are building this infrastructure, though production-ready shared sequencing remains an open challenge.

L2 Security Considerations

Sequencer Centralization: Over 90% of L2s still use a single permissioned sequencer. Users must trust the sequencer not to censor transactions or extract excessive MEV.

Bridge Security: L2 bridges remain the highest-risk attack surface. Users should prefer canonical bridges and understand the trust assumptions of third-party bridges.

Data Availability Risk: L2s using off-chain data availability (validiums) introduce additional trust assumptions. Users should verify the data availability model of each L2 they use.

Stage Classification: L2Beat classifies L2s by decentralization stage. Stage 1 requires functioning fraud/validity proofs. Stage 2 requires permissionless validation. Most major L2s are at Stage 1 in 2026.

Future Trends

L2 Consolidation: The market is consolidating around Base, Arbitrum, and Optimism. Smaller L2s face existential challenges without strong distribution or technological differentiation.

ZK Prove Maturation: Advances in proving systems (Stwo, plonky3) continue to reduce proof generation costs, narrowing the gap between optimistic and ZK rollups.

Chain Abstraction: Users increasingly interact with applications without knowing which L2 they’re using, thanks to account abstraction and cross-chain messaging standards.

Institutional L2s: Exchange-backed L2s (Base, Ink, Mantle) and traditional finance partnerships are driving institutional adoption of L2 infrastructure for tokenized assets and settlement.

Security Considerations

Rollup-Specific Risks

While Layer 2 solutions inherit security from Ethereum, they introduce additional risks specific to their architectures. L2Beat’s stage classification provides a useful framework — networks are ranked Stage 0 (training wheels), Stage 1 (limited fraud proof system), or Stage 2 (fully permissionless). As of 2026, most major L2s have achieved Stage 1, but few have reached Stage 2.

Sequencer risks include potential censorship, transaction reordering, and MEV extraction beyond what occurs on Layer 1. Over 90% of L2s still operate permissioned sequencers, creating a trust assumption that users must evaluate. The Espresso shared-sequencer network continues development as a potential solution.

Data availability risks emerge when data is stored off-chain. The Celestia and EigenDA ecosystems provide alternative DA layers that some L2s use to reduce costs while maintaining stronger guarantees than fully centralized alternatives.

Smart contract risks remain significant. Layer 2 bridge contracts and protocol infrastructure are complex, and exploits continue to occur — though the industry has matured in its response. The Arbitrum Security Council’s ability to freeze 30,766 ETH linked to the KelpDAO hack in April 2026 demonstrates both the power and the centralization risk of these governance mechanisms.

L2 Consolidation Risk

The biggest market-level risk in 2026 is chain death. Of the dozens of L2s launched over the past three years, 21Shares estimates most will not survive past 2026. Networks without sustainable economic models, strong distribution, or differentiated technology face zombie chain status — still technically live but with negligible usage, liquidity, and development activity. Blast’s TVL collapse from $2.2 billion to $55 million exemplifies this risk.

Mitigation Strategies

Users can mitigate risks by distributing assets across multiple L2 networks, using canonical bridges where possible, and evaluating a network’s L2Beat stage classification. For developers, designing for multi-chain deployment and minimizing dependence on any single L2’s uptime or liquidity is essential risk management.

Staying informed about security incidents, protocol upgrades, and governance changes is critical in an ecosystem that evolves rapidly. Resources like L2Beat provide real-time monitoring of risk parameters across the L2 landscape.

The Future of Layer 2

Consolidation and Specialization

The L2 ecosystem is undergoing a brutal consolidation. Three networks — Base, Arbitrum, and Optimism — now process nearly 90% of transactions. 21Shares predicts most smaller L2s will not survive past 2026. The survivors will be those with strong distribution (exchange-backed networks like Base, Mantle, Ink), ETH-aligned designs, or genuine technological differentiation.

Simultaneously, the ecosystem is evolving toward specialization. While early rollups attempted to serve all use cases, networks are increasingly optimized for specific applications — gaming chains, DeFi chains, identity chains — with interoperability mechanisms connecting them. MegaETH represents the high-performance thesis: a real-time blockchain targeting 100,000 TPS with 10-millisecond block times.

Chain Abstraction

The end state of L2 proliferation is abstraction. Users will interact with applications without needing to know — or care about — which chain they’re using. Account abstraction already enables smart contract wallets that interact seamlessly across any L2. Chain abstraction frameworks provide developers simple APIs for cross-chain functionality, hiding bridge integration and address translation complexity.

Institutional Adoption

The biggest growth vector for L2s in 2026 is institutional. Tokenized real-world assets, regulated settlement networks, and enterprise blockchain deployments are choosing L2 infrastructure for its combination of Ethereum-grade security and low transaction costs. Cari Network, the first multi-bank tokenized deposit network built on ZKsync, exemplifies this trend — five founding banks with over $600 billion in deposits exploring on-chain settlement that reduces T+2 cycles to seconds.

Robinhood building on the Arbitrum stack for tokenized equity trading similarly signals mainstream financial infrastructure migrating to L2s. Base capturing 43% of on-chain BTC spot volume demonstrates that even Bitcoin-centric activity is channeling through Ethereum L2 rails.

Conclusion

Layer 2 solutions have transformed Ethereum from a network struggling with congestion into a scalable platform capable of supporting global adoption. By 2026, the L2 ecosystem holds over $34 billion in value, processes millions of transactions daily at sub-cent fees, and serves as the primary interface for most blockchain users.

The market has consolidated around a small number of winners — Base for consumer scale and distribution, Arbitrum for DeFi depth and liquidity, Optimism for the Superchain vision, and specialized ZK rollups like zkSync and Starknet for applications demanding cryptographic finality. The choice between these involves genuine tradeoffs, and there is no single optimal solution.

The defining trend of 2026 is the shift from experimental infrastructure to production-grade platform economics. L2s are no longer judged by fee reduction alone but by developer stack quality, liquidity depth, institutional readiness, and sustainable revenue models. The winners are building toward chain abstraction — a future where users transact on Ethereum’s full security without ever needing to know which L2 they are using.

L2 Use Cases by Category

How to Choose an L2 for Your Project

Security Requirements: If immediate finality is critical (e.g., settlement, high-value transfers), ZK rollups like zkSync Era or Starknet provide cryptographic guarantees without challenge periods.

Ecosystem Maturity: For DeFi projects requiring deep liquidity and composability with existing protocols, Arbitrum offers the most mature ecosystem with established lending, DEX, and derivatives protocols.

User Acquisition: For consumer-facing applications targeting retail users, Base provides the largest user base and seamless Coinbase onboarding.

Customization Needs: Projects requiring custom gas tokens, block parameters, or governance models should consider Arbitrum Orbit or OP Stack based L3 solutions.

Future-Proofing: For applications planning to operate across multiple L2s, Optimism’s Superchain and ZKsync’s Elastic Network offer native interoperability roadmaps.

Performance Benchmarks 2026

L2 Interoperability Protocols

Superchain: Optimism’s ecosystem of OP Stack chains with shared bridge, security, and governance. Native interop enables atomic composability between Superchain L2s.

Arbitrum AnyTrust: Framework for launching custom L2/L3 chains with configurable trust assumptions. Orbit chains can communicate through the AnyTrust protocol.

Elastic Network: ZKsync’s vision of interconnected ZK chains sharing proving infrastructure and liquidity through the ZKsync Gateway.

Polygon AggLayer: Aggregation layer connecting Polygon chains and other L2s through ZK proofs, enabling unified liquidity and cross-chain execution.

L2 Developer Toolkit

Essential tools for building on Layer 2 in 2026:

• Hardhat/Foundry: Smart contract development and testing - works on any EVM L2
• Arbitrum SDK: Chain-specific APIs for Orbit, AnyTrust, and custom gas tokens
• OP Stack CLI: Deploy and manage OP Stack chains with standardized tooling
• L2Beat: Monitor L2 security parameters, TVL, and decentralization status
• Cross-L2 Explorers: Etherscan-style explorers for all major L2s with unified search
• Bridge Analytics: Track bridge flows, TVL, and activity across L2s with Dune Analytics

Resources

• Arbitrum Documentation
• Base Documentation
• Optimism Documentation
• zkSync Era Documentation
• Starknet Documentation
• L2Beat - Layer 2 Monitoring
• L2Beat - Risk Classification
• The Block - 2026 Layer 2 Outlook
• Superchain Eco