OP Stack vs. Arbitrum Orbit: The Best L2 Rollup Comparison

Previous articles in this series:

Gelato's Guide to Avalanche L1s and Native Interoperability

L1 Blockchain Stacks: Avalanche vs Cosmos

Rollup L1s: How ABC Stack's Celestia Sovereign Rollup L1s Compare to Avalanche and Cosmos L1 blockchains

Tl;DR

• Arbitrum and Optimism are leading Layer 2 scaling solutions for Ethereum, both optimistic rollups with 7-day challenge periods, but with different technical implementations

• OP Stack (Optimism) focuses on Ethereum alignment with strict EVM equivalence, a Superchain architecture for interchain communication, and a permissionless bisection protocol for fraud proofs using Cannon FPVM

• Arbitrum Orbit offers more flexibility with its Nitro stack, supporting both traditional Ethereum rollups and AnyTrust mode (using a Data Availability Committee), along with Stylus VM for non-EVM languages like Rust and C++, and a Universal Intents Engine that enables fast cross-chain operations through competitive solver execution.

• Arbitrum emphasizes custom stack innovations that diverge from strict Ethereum equivalence—such as its AnyTrust DAC, fast withdrawal mechanisms, and Layer 3 strategy. In contrast, the OP Stack focuses on standardizing Ethereum-aligned practices, prioritizing equivalence and interoperability across its L2 Superchain ecosystem.

• OP Stack is MIT open-source, with fees only applying when joining the Superchain (2.5% of revenue or 15% of onchain profit). Arbitrum uses a Business Source License with an 'Additional Use Grant' allowing free L3s on Arbitrum One while requiring a 10% profit share for independent chains. Both systems have different approaches to revenue sharing, with OP focusing on collective governance while Arbitrum emphasizes 'Your chain, your rules' flexibility.

• OP-Succinct Lite and Arbitrum's BoLD protocol represent attempts to improve the traditional fraud proof mechanism, with the former using ZK proofs to reduce withdrawal times to 1 day.

Introduction

Arbitrum and Optimism provide optimistic Layer 2 scaling solutions for Ethereum that improve transaction speed and reduce costs while aiming to extend Ethereum's security. While they share foundational traits such as Ethereum settlement and data availability/consensus, optimistic fraud proofs, and EVM compatibility, their technical trajectories reveal tradeoffs between aligning with Ethereum and pursuing chain-specific optimizations. Looking ahead, ZK technology and renewed Ethereum-centricity will likely transform these stacks significantly. Both platforms remain works in progress, with Optimism's Superchain, crucial for its interoperability vision, not yet launched.

What is an optimistic rollup?

State transitions are initially “optimistically” accepted without proofs but remain subject to a challenge period (7 days on mainnet). During this window, anyone can challenge invalid state transitions through a fault proof mechanism. Both Optimism and Arbitrum implement this model with their own technical approaches and tradeoffs.

How L2 settlement happens

For optimistic L2s, Ethereum settlement specifically means having a smart contract light client bridge deployed on Ethereum that receives:

1. Data availability (DA) batches containing compressed transaction data
2. State root commitments that represent the updated L2 state
3. Fraud proofs that ensure the correctness of state transitions

This smart contract bridge secures canonical asset transfers between L1 and L2. This direct connection to Ethereum's security model is what constitutes Ethereum settlement.

Data availability and EVM compatibility

The OP Stack publishes compressed transaction batches to Ethereum to ensure data availability and security. The op-batcher aggregates sequencer batches into "channels," which are sequences of Layer 2 blocks, to optimize compression and reduce data availability costs.

The EVM module in the OP Stack is a lightly modified version of the EVM that adds support for L2 transactions initiated on Ethereum and incorporates an L1 Data Fee. This fee accounts for the cost of publishing transaction data to Ethereum, ensuring data availability and security for L2 transactions. The execution layer of the OP Stack, which processes these transactions, is implemented using op-geth, a fork of the go-ethereum client.

OP Fault Proof System

The Fault Proof System is a decentralized security mechanism designed to ensure the validity of state transitions and enable trustless withdrawals from Optimism to Ethereum. There are three components of OP stack’s fault proof system:

• Fault Proof Program (FPP): Verifies L2 state transitions by replaying transactions using L1 input data. It combines consensus and execution logic to ensure deterministic outcomes.

• Fault Proof Virtual Machine (FPVM): Executes the FPP in a sandboxed environment, enabling onchain dispute resolution. Different VMs can be used for verification.

• Dispute Game Protocol: A permissionless system where users challenge invalid transactions by submitting bonds. Honest challengers are rewarded, while malicious actors lose their stake.

Cannon is Optimism's default Fault Proof Virtual Machine (FPVM) that verifies state transitions. It implements a virtual machine that can execute MIPS instructions and generate cryptographic proofs of execution that can be verified within the EVM. MIPS is an instruction set that helps processors run programs faster and more reliably, which is why it’s suitable for this kind of verifiable computation.

It has two main parts: one in Solidity that verifies MIPS instructions onchain and an offchain Go implementation (mipsevm) for generating proofs. Most computation happens offchain, only requiring minimal verification steps to be performed onchain, which is crucial for cost-effectiveness given Ethereum's gas limitations.

The Cannon dispute resolution process goes as follows:

1. Dispute: Participants submit competing claims about the L2 state by posting bonds.

2. Bisection: The disputed computation undergoes multiple rounds of bisection, where participants take turns challenging specific portions of the execution trace. This requires multiple on-chain transactions as the dispute narrows down from entire state transitions to specific instruction disagreements.

3. Generate Proof: Only after many rounds of bisection and once a single instruction is isolated does Cannon create a witness proof for that specific step.

4. Verify Onchain: MIPS verifies just this single instruction, the culmination of the multi-round bisection process.

5. Resolve: The dispute game concludes by awarding the bond to the honest party based on the verified instruction outcome.

This allows anyone to challenge invalid state transitions while ensuring that the costs of verification remain practical.

Optimism Superchain

The Optimism Superchain is a unified network of interconnected L2 blockchains (OP Chains) built using the modular OP Stack. Superchain interoperability enables chains to read each other's state while maintaining security. It scales by adding more interoperable L2 chains rather than upgrading a single one.

Superchain achieves interoperability through several mechanisms. Cross-chain messaging is facilitated by the OP Supervisor, which maintains a database of log events across chains, validates these events, and ensures transaction safety by referencing Ethereum's consensus layer. Importantly, the OP Superchain interoperability features are not yet live in production.

Source: docs.optimism.io/stack/interop/explainer

The Superchain implements a fully connected mesh network where each blockchain has direct dependencies with every other blockchain in the cluster, allowing any chain to transact directly with any other chain. Each blockchain in the Superchain shares the same security model and governance process.

The Superchain introduces strong standardization requirements, which means some experimental OP Stack features may be excluded or delayed, becoming available only after a thorough review and approval process.

OP Stack special features

What are OP Flashblocks

Even 2-second block times on modern rollups like OP Stack L2s are too slow for applications like trading, gaming, and payments that need web2-like (<100ms) response times. Flashblocks act like regular blocks but don't require state root computation or header storage, significantly reducing overhead while maintaining the same security properties. Through Rollup Boost, Optimism Stack chains, can implement Flashblocks to achieve ultra-low latency transaction confirmations.

Removal of Custom Gas Tokens in Optimism

All new OP Stack chains must use ETH as the gas token. Existing chains that use custom gas tokens may either maintain their own fork (without official support or future compatibility) or migrate to the standard ETH gas token configuration. Although previously supported, custom gas tokens implemented as a change at the protocol level are unnecessary after Account Abstraction allows users to pay gas in any token at the application layer.

What is OP altDA

Alt-DA mode allows for OP Stack chains to integrate with various DA Layers beyond Ethereum. It’s aimed to address the challenge of increasing EIP-4844 blog fees due to congestion on Ethereum. This however causes issues with trust assumptions where chain operators must be relied upon to post both correct data to the DA Layer and valid commitments to Ethereum L1. Optimism’s move to being more Ethereum-centric, makes this feature more redundant.

What are OP-Succinct ZK Fraud Proofs

Succinct and Optimism introduced OP Succinct Lite in late February, effectively trying to replace the traditional multi-step interactive fraud proof mechanism with ZKPs. With OP Succinct Lite, zero-knowledge proofs are used to resolve disputes in the OP Stack rollup system.

What are ZK fraud proofs?

Initially, validity proofs proactively verified every transaction batch upfront. Now, a ZK fraud proof (used in hybrid rollups) resolves disputes by reactively generating zero-knowledge proofs when transactions are challenged. This reduces withdrawal times compared to traditional optimistic rollups, slashing finality times to 1 day.

How OP Succinct Lite works in a nutshell

OP Succinct Lite is a system that uses zero-knowledge proofs specifically for dispute resolution in optimistic rollups. This is to simplify the challenge process while maintaining the optimistic execution model.

1. Transactions are still being processed offchain and assumed valid until challenged (optimistic)
2. ZK proofs are only used/generated when disputes arise (rather than for every transaction) to cryptographically verify the validity of the disputed state or transaction
3. The dispute is resolved through onchain verification of these proofs submitted to an L1 smart contract

As we can observe, it serves as a middle ground between traditional interactive fraud proofs and full validity proofs (full ZK rollups).

Similar to OP Succinct Lite, Kailua, developed by RISC Zero late last year, is a hybrid rollup framework that uses ZKPs only when disputes arise in optimistic rollups. By leveraging RISC Zero’s zkVM, Kailua enables fast, single-round fraud proofs and reduces finality times to about one hour.

Neither Succinct nor Kailua is currently included in the official Superchain stack standards.

What is Arbitrum Orbit and How Orbit Rollup works

The Orbit stack utilizes the Nitro stack for its foundation. Arbitrum Orbit allows developers to deploy customizable chains using the Nitro stack. These chains can be configured as Layer 2 (L2) rollups settling on Ethereum or Layer 3 (L3) chains settling on any Ethereum L2, such as Arbitrum One.

It utilizes the rollup data availability model, posting all transaction data directly to the parent chain (Ethereum for L2, or another L2 for L3) rather than using a Data Availability Committee (DAC).

Data availability and EVM compatibility

Arbitrum offers two operational modes for data availability: Rollup and AnyTrust. In Rollup mode, all transaction data is posted directly to Ethereum, either as calldata or via EIP-4844 blobs. This is to ensure that the chain inherits Ethereum’s security guarantees and that anyone can reconstruct the chain’s state from on-chain data. AnyTrust mode, however, leverages a DAC and only posts cryptographic proofs to Ethereum. This significantly reduces costs but introduces a trust assumption (at least one committee member will behave honestly and make the data available) regarding the committee’s honesty.

Nitro is Arbitrum's execution environment, and like Optimism, is built on a fork of go-ethereum with specific modifications for L2 functionality. It separates execution from proving by using Go's native compiler for high-performance transaction processing while compiling to WebAssembly for deterministic fraud-proof verification. Nitro supports EIP-4844 blob transactions for further cost reduction and implements a State Transition Function (STF) that deterministically advances chain state based on sequenced messages. This guarantees security and allows for objective fraud proof resolution, all without requiring consensus between L2 nodes as all Arbitrum Nitro nodes process the same ordered list of transactions and arrive at the same state independently.

Arbitrum's Stylus introduces WASM support for non-EVM languages. It allows developers to write smart contracts in languages that compile to WebAssembly (WASM), such as Rust, C, and C++. Stylus contracts are fully interoperable with Solidity contracts, meaning they can call each other. Stylus’ ability to leverage optimized WASM-compatible languages allows it to deliver dramatically lower gas fees and faster execution.

Arbitrum’s Fraud Proof System

Arbitrum has a multi-round fraud proof system that resolves disputes through an interactive "dissection" process instead of re-executing entire transactions on Ethereum.

This reduces gas costs compared to single-round systems (such as in Optimism) as it minimizes the computational burden placed on Ethereum's Layer 1 during dispute resolution. However, this doesn’t mean shorter finality because the optimistic security model still requires a challenge period during which any validator can dispute a transaction, regardless of how efficient the fraud proof mechanism is. The BoLD protocol aims to mitigate delays by bounding dispute resolution timelines, but transactions still face a near 7-day withdrawal window when bridging to Ethereum. Just like with Optimism, it creates issues particularly for time-sensitive applications.

Arbitrum’s recent adoption of the BoLD protocol also removes the need for permissioned validators, allowing anyone to participate in the fraud proof process. This change improves decentralization and aligns Arbitrum’s security model more closely with Ethereum’s. BoLD’s interactive dispute resolution and one-step proof mechanism ensure that only the specific disputed computation is executed on Ethereum, minimizing gas costs. However, despite these improvements, the withdrawal period for bridging assets, typically around 7 days, remains in place to maintain security, which can still be a limitation for certain applications.

How Arbitrum AnyTrust works

AnyTrust allows chains to opt for a Data Availability Committee (DAC) instead of Ethereum’s full data availability. It uses a permissioned DAC (predefined set of explicitly selected or approved trusted entities) to validate transaction data offchain, reducing costs by minimizing onchain data posting. Importantly, when using AnyTrust, a chain is no longer strictly an Ethereum rollup, as it significantly reduces the extent to which it leverages Ethereum's security model.

It relies on at least two honest committee members for data integrity. If the DAC cannot provide sufficient signatures, the sequencer falls back to posting the full data directly to Ethereum, ensuring continued data availability. Chains like Arbitrum Nova leverage this model for ultra-low fees while maintaining EVM compatibility, allowing developers to choose between full Ethereum data availability or DAC-based solutions based on cost and decentralization needs.

Fast Withdrawals for Anytrust Chains

Instead of the 7-day challenge period required by Optimistic Rollups, transactions on AnyTrust chains with fast withdrawals will be processed by a committee of validators. The unanimous vote of the validator committee replaces the waiting period, as a result achieving faster finality. Fast withdrawals can reduce the withdrawal period from 7 days to as little as 15 minutes.

Fraud proofs are no longer the primary security mechanism when fast withdrawals are enabled. Fast Withdrawal mode essentially circumvents the traditional fraud proof mechanism by adding a trusted validator committee that can finalize transactions before the challenge period completes. Since AnyTrust chains already have a trust assumption placed on their DAC, DAC members can also serve as validators to avoid introducing new trusted parties.

Arbitrum L3s

An Arbitrum L3 chain is a chain that settles to an L2 chain instead of directly to Ethereum (L1). It vertically scales by stacking purpose-specific rollups over an L2 base. So we have:

• L1: Ethereum mainnet - the base blockchain
• L2: Chains like Arbitrum One or Arbitrum Nova that settle canonical assets on Ethereum
• L3: Orbit chains that settle to an L2 chain (like Arbitrum One) rather than to Ethereum

An Orbit L3 chain uses an L2 chain (such as Arbitrum One) as its parent/settlement layer instead of using Ethereum directly. This creates a chain hierarchy where Ethereum provides the foundational security and an L2 like Arbitrum One builds on top of Ethereum. Finally you would have an L3 Orbit chain built on top of that L2.

It's important to understand that L3s are essentially like L2s but with a different "L1”. In this case, Arbitrum One becomes the settlement layer for the L3. This creates a fundamental difference in security properties:

• L2 on Ethereum: Directly inherits Ethereum's data availability, consensus, AND settlement
• L3 on Arbitrum: Inherits Ethereum's data availability and consensus, but NOT direct Ethereum settlement

This means that while an L3 on Arbitrum One still uses Ethereum for DA & consensus, it does not have direct settlement on Ethereum. Users cannot directly force-exit their assets from the L3 to Ethereum using the standard L2 bridge mechanism that's fundamental to Ethereum's L2 security model.

L3 Chains Settle to Arbitrum, Not Ethereum

As mentioned earlier, Orbit’s L3 chains settle transactions on L2 chains (such as Arbitrum One or Base) which in turn settle on Ethereum. This results in a weaker security inheritance chain as you’re not directly settling on Ethereum, and the finality of L3 transactions depends on the security and uptime of both the L2 and L1.

Arbitrum Orbit expansion outside of Ethereum

Arbitrum is extending its reach beyond the Ethereum ecosystem. Orbit L2s are already launching, or rumored to launch, on alternative L1s like TON (Duck Chain), Berachain (Pretzel Layer), and HyperEVM. Arbitrum encourages developers across all chains to adapt the Nitro stack for seamless Orbit L2 deployments on their preferred L1s.

Intent-based interop with Intents Engine

Unlike traditional transactions that define exact steps, intents express what you want accomplished, allowing specialized entities called "solvers" to compete to find the most efficient execution path. Intents are user-declared desired outcomes (like "swap token A for token B" or "move assets from chain X to chain Y") without specifying the exact execution path.

Arbitrum's Universal Intents Engine is a universal system for processing cross-chain requests with four key components:

• A standardized message format for expressing intents

• Broadcast contracts for secure cross-chain communication

• Fast settlement solutions to reduce execution time

• An intent dissemination feed where users publish requests for solvers to execute

Intent-based interoperability using Arbitrum's engine allows assets and data to move seamlessly between different blockchains without users needing to understand the complex underlying mechanics. A user simply expresses what they want (e.g., "transfer USDC from Arbitrum One to Ethereum"), and the intent system broadcasts this to competing solvers who find the fastest, cheapest execution path and complete the operation in under three seconds while maintaining trustlessness throughout the process.

The Universal Intent Engine is set to launch by Q1 2025, with native crosschain operations rolling out in Q3 2025. It’s in active development, similar to OP’s Superchain native interop.

How OP stack vs. Orbit differs

Ethereum scaling solutions must reconcile security with flexibility, often forcing tradeoffs between decentralized verification and customizable chain logic. OP Stack (Optimism) and Orbit (Arbitrum) address this with distinct technical architectures.

On Fraud Proof Systems

OP Stack uses optimistic rollups with a permissionless bisection protocol for fraud proofs, narrowing disputes to single instruction steps verified on-chain via the Cannon fault-proof virtual machine (FPVM). It operates with a 7-day challenge window and economic bonding to incentivize honest behavior.

Orbit relies on Arbitrum Nitro’s interactive fraud proofs, resolving disputes through multi-round challenge games. Additionally, Orbit supports AnyTrust chains, which utilize data availability committees (DACs) to expedite finality and allow developers to configure trust assumptions.

During disputes, both chains avoid full transaction re-execution. Optimism isolates MIPS instructions while Arbitrum narrows to WASM opcodes. Though Arbitrum’s granular WASM-level disputes remain slightly more gas-efficient for complex computations. The 7-day withdrawal window persists in both systems due to their shared optimistic security model.

On Licensing and Fees

OP Stack is completely open-source with an MIT license, with fees only applying if you join the Superchain. While developers can technically deploy chains without joining the Superchain, the official OP team primarily promotes Superchain usage, so most OP Stack deployments are expected to join and contribute to the collective.

Under the Law of Chains, OP Chains within the Optimism Superchain adhere to a standardized revenue-sharing model. Each chain contributes the greater of:

• 2.5% of the gross Sequencer Revenue

• 15% multiplied by an amount equal to (x) the gross Sequencer Revenue less (y) the L1 Gas Fees

This ensures that all participating chains support the Optimism Collective’s mission to fund public goods and ecosystem development.

Arbitrum Orbit uses a Business Source License (BSL) with an “Additional Use Grant,” similar to the models adopted by Uniswap and Aave. This setup allows projects to launch L3 chains that settle to Arbitrum One or Nova without needing a license or sharing revenue with the Arbitrum DAO. However, chains that choose to settle elsewhere, like Ethereum L2s, must obtain a license and share a portion of sequencer profits with the DAO, typically 10%.

This approach strikes a balance between the restrictive “Law of Chains” that governs the OP Superchain and the open-source flexibility of the OP Stack. It offers flexible data availability, extensive customization, and sovereign governance, aligning with the ethos of “Your chain, your rules.”

Node Infrastructure

Arbitrum Nitro is the node software and architecture currently powering Arbitrum. It uses a WASM-based node design, compiling Geth EVM to native code for fast execution and switching to a WebAssembly variant (WAVM) only when running fraud proofs. This setup lets validators run efficiently most of the time, with fraud proofs isolating only the disputed code, which reduces on-chain costs during challenges.

Optimism Bedrock features a modular node architecture that separates execution, consensus, and data derivation. Its execution runs on op-geth, a lightly modified Ethereum client for EVM equivalence and features like L1 data fee accounting. Most computation occurs offchain, minimizing onchain gas costs.

On VM Architecture

Optimism maintains strict EVM equivalence via op-geth, prioritizing compatibility over execution optimizations. It preserves L2 status through strict EVM compliance. OP Stack processes transactions in a single, unified way and settles directly on Ethereum, while Orbit uses interactive fraud proofs and allows L3 chains to settle on Arbitrum’s own networks instead of Ethereum itself, which means Orbit’s approach doesn’t fully follow Ethereum’s security model.

Arbitrum’s Stylus, a parallel WASM-based VM that runs alongside the EVM to support Rust/C++ contracts, creates a hybrid execution environment that’s incompatible with Ethereum’s Native Rollup vision due to WASM’s inability to natively verify via Ethereum’s EVM. Orbit’s multi-VM approach and AnyTrust’s reliance on centralized data committees position its chains as application-specific scaling layers rather than true Ethereum L2s.

Which to choose and why?

When evaluating Optimism's OP Stack and Arbitrum's Orbit, it becomes clear that each solution caters effectively to distinct segments of the Layer-2 market. OP Stack prioritizes Ethereum alignment, delivering standardized, EVM-equivalent rollups tightly integrated with Ethereum's security and decentralization. Arbitrum Orbit, while fully capable of being deployed as Ethereum Rollup L2s, has strategically positioned itself differently, leveraging greater customization to achieve higher performance and substantial data availability (DA) cost efficiencies through its AnyTrust Data Availability Committees (DACs). This tailored approach has clearly resonated with the market, as evidenced by approximately half of the Orbit chains today adopting the AnyTrust model, highlighting a significant user preference for flexibility over strict Ethereum alignment.

An often underestimated factor in choosing between OP Stack and Arbitrum Orbit is their licensing and fee structure. OP Stack is fully open-source under an MIT license, with fees only applying to chains that join the Superchain, contributing a standardized cut (greater of 2.5% gross sequencer revenue or 15% of sequencer revenue − L1 gas fees) to the Optimism Collective. In contrast, Arbitrum Orbit uses a Business Source License, allowing free L3 deployments on Arbitrum One or Nova but requiring a 10% sequencer profit share for chains settling elsewhere. This highlights a key philosophical divide: Optimism promotes collective governance and Ethereum alignment, while Arbitrum favors sovereignty and flexibility. For developers, the decision often hinges on ecosystem integration versus operational autonomy.

Moreover, Arbitrum's introduction of Stylus exemplifies a robust, developer-oriented innovation by expanding smart contract programming beyond traditional EVM languages. While this innovation has substantially broadened developer accessibility and performance capabilities, it does raise potential compatibility challenges for Orbit chains aspiring to become Ethereum-native rollups in the future. Early research into native rollups emphasizes the necessity of strict EVM equivalence, suggesting that Orbit's more permissive environment could limit seamless Ethereum-native integration down the road.

OP has taken an early lead in native interoperability with its Superchain framework, attracting high-profile projects like Base, Worldchain, Lisk, and Ink. This alignment showcases a strong conceptual win for OP, positioning it as a potential standard for an interconnected L2 ecosystem. However, the rollout is still in its early stages, and practical benefits remain untested, with some early signals suggesting that not all partners, like Base, may join the initial native interop set.

Meanwhile, Arbitrum is pushing forward with its own approach, the Universal Intents Engine, which aims to deliver flexible, cross-chain intent-based interoperability. With both models still evolving, it’s too soon to predict which approach will ultimately offer greater impact.

While both stacks have distinct approaches to fraud proofs, these differences may soon become less relevant. Current fraud proof systems are likely transitional technologies that will eventually be replaced by ZK variants offering faster finality and enhanced security.

It's important to recognize that both Optimism and Arbitrum represent early iterations of L2 technology that will continue to evolve. Neither stack should be seen as finalized, as both are likely to evolve significantly with ongoing advancements in zero-knowledge (ZK) technology and a potential return toward more Ethereum-centric design principles. New approaches like Ethereum Native Rollups and Based Rollups aim to reinforce Ethereum's position in the L2 architecture. These developments raise fundamental questions about what truly constitutes an L2 and how Ethereum’s security is meaningfully inherited, topics we’ll explore in our forthcoming articles.

For developers looking to integrate their applications with Gelato Web3 Services, check out Web3 Functions, Relay, and VRF! Visit our Discord server for developer support and engagement, and stay updated with the latest developments by following us on X.

Definitions

• Optimistic Rollup: A scaling solution that processes transactions off-chain and posts transaction data to Ethereum, assuming transactions are valid by default but allowing a challenge period where fraud proofs can be submitted.

• Fraud Proof: A mechanism that allows participants to challenge and prove the invalidity of state transitions in optimistic rollups, securing the system against malicious behavior.

• ZK Fraud Proofs (as used by systems like Succinct): A hybrid approach that combines zero-knowledge proofs with fraud detection. Instead of proactively proving every transaction, a ZK proof is generated only when a dispute is raised, proving the incorrectness of a specific claim. This enables faster fraud resolution with lower on-chain costs, while maintaining the challenge-based model of optimistic rollups.

• Validity Proofs (ZK Rollups): A model where every batch of transactions is accompanied by a cryptographic proof (e.g., SNARK or STARK) that verifies the correctness of state transitions. These proofs are verified on-chain, enabling instant finality and eliminating the need for a challenge period, but requiring more upfront computation and prover infrastructure.

• Data Availability (DA): Refers to how and where transaction data is published so that it can be independently verified by others. In Ethereum L2 rollups, DA typically involves posting transaction data on Ethereum (e.g., calldata), ensuring that anyone can reconstruct the L2 state. Alternative models, like DACs or off-chain solutions, reduce costs but weaken Ethereum-based security assumptions.

• Data Availability Committee (DAC): A permissioned group of trusted entities responsible for storing and serving transaction data off-chain, used as an alternative to posting full data on Ethereum. While it reduces costs, it introduces trust assumptions, as data availability depends on the honesty and liveness of the committee members.

• Settlement: The process by which an L2 anchors its state to Ethereum, including enforcing state transitions and finalizing disputes (e.g., through fraud or validity proofs). True Ethereum settlement implies that the L2 ultimately relies on Ethereum’s consensus, finality, and dispute resolution, rather than third-party trust assumptions.