Layer2 clusters launch Stacks to seize developer resources. Who will be the final winner?

Author: Grapefruit, ChainCatcher

Since Optimism announced the launch of the OP Stack for building OP Rollup chains and proposed the concept of "Superchain," Arbitrum has announced the launch of Arbitrum Orbit, focusing on expanding the L3 market, zkSync has created the ZK Stack for developing ZK Rollup chains, Starknet has released the Starknet Stack toolkit for building application chains (Appchains), and Polygon has launched the Polygon CDK for developing zkEVM chains.

The so-called Stack refers to the standardized, componentized, and modular technical architecture or open strategies and SDKs made by developers. Based on such component architectures or strategies, one can quickly build or launch their own blockchain network.

From the perspective of product development quantity and trends, the current L2 Stacks have evolved from the Layer2 market into an independent track, shifting the competition between Layer2 networks from the initial focus on the number, variety, and prosperity of on-chain ecological applications to the competition over the size of the Superchain kingdom.

This is because, after various Layer2 networks announced their own Superchain network solutions L2 Stacks, the original single-chain gradually developed into a more open platform or infrastructure product, focusing on developing the Superchain ecosystem. The narrative of competition in the Layer2 track has also changed; the competition between Layer2 networks is no longer just about the quantity of on-chain applications but rather about how many chains are built based on their Stacks or how many developers or project parties choose to use their component tools when building chains.

Currently, mainstream Layer2 networks have launched corresponding Stack solutions, mainly including OP Stack, Arbitrum Orbit, Polygon CDK, ZK Stack, and Starknet Stack.

So, what are the differences between these L2 Stacks? Faced with numerous "one-click chain launch" tools, how will developers choose? How can one evaluate whether a solution is suitable for themselves?

L2 Stacks: The Two Camps of Optimistic Rollup and ZK Rollup

From a timeline perspective, Optimism introduced the OP Stack concept in October 2022, making it the earliest Layer2 network to propose L2 Stacks products. This was followed by Arbitrum's launch of Arbitrum Orbit in March this year, zkSync's introduction of the ZK Stack for building ZK Rollup chains in June, StarkNet's announcement of the Starknet Stack toolkit for customizing Appchains at the Paris EthCC conference in July, and Polygon's launch of the Polygon CDK for developers to build zkEVM L2 networks on August 31. Therefore, in the L2 Stacks track, OP Stack is a pioneer.

Based on the different consensus algorithms used by the underlying networks, the mainstream L2 Stacks solutions in the current market can be mainly divided into two camps: the Optimistic Rollup camp (abbreviated as OP Rollup) and the ZK Rollup camp. The former is represented by OP Stack and Arbitrum Orbit, while the latter mainly includes Polygon CDK, ZK Stack, and Starknet Stack.

In terms of product operation logic, L2 networks built on the OP Rollup camp's Stack operate similarly to Optimism (now the OP Mainnet), following the process of "off-chain (L2) handling the actual computation and storage of smart contracts, on-chain (L1) verifying and storing transaction data, using the consensus algorithm of optimistic fraud proofs." In contrast, L2 networks built on the ZK Rollup camp's Stack, like the current zkSync and Polygon zkEVM mechanisms, use zero-knowledge proof algorithms for submitting data from L2 to L1.

In short, a newly built L2 network based on which Layer2 Stack solution possesses the advantages of that Layer2 network; one can even view these newly built L2 Stacks networks as homogeneous competitors of their networks. For example, an L2 network built on the OP Stack has the same advantages as the Optimism network.

So, what are the differences between the same camp's L2 Stacks? The main differences lie in the openness of each Stack and their expansion strategies.

OP Camp: OP Stack Expands Ecosystem Through Collaboration, Arbitrum Orbit Focuses on Enhancing Developer Experience

*

#### Leader OP Stack Expands Superchain Ecosystem Through External Collaboration

OP Stack is a standardized, modular open-source toolkit launched by Optimism, mainly comprising the data availability layer, execution layer, settlement layer, governance layer, and other modules and tools needed to build a blockchain. Based on these components, developers can customize their own Layer2 networks using OP Stack according to their needs.

Optimism plans to build a Superchain empire based on OP Stack, where the so-called Superchain refers to the Layer2 networks built using OP Stack collectively (also known as OP chains).

The concept of the OP Stack Superchain differs from the multi-chain concepts of Cosmos and Polkadot. The underlying architecture of the Superchain is uniformly standardized, with a consistent consensus algorithm, shared security, and ultimately secured by the Layer1 blockchain (Ethereum mainnet). Additionally, due to the uniform standardized architecture of the internal chains, resources and information between chains can be shared through the Optimism Bridge. In contrast, the multi-chain architectures of Cosmos and Polkadot introduce new consensus algorithms on each chain, requiring developers to start new validator sets whenever a new chain is activated. Moreover, the security and information between chains cannot be directly shared; cross-chain information needs to be relayed through a hub chain or IBC protocol.

In the OP Stack Superchain ecosystem, users do not need to focus on whether it is Layer2 chain A, B, or C; they can view it as a whole "Superchain." Developers building applications also target the entire Superchain. For example, within the Superchain ecosystem, only one RPC node is needed to support users in trading on any OP chain without switching networks.

Currently, the OP Stack Superchain kingdom mainly consists of the officially launched OP Mainnet and several Layer2 networks built on OP Stack.

In terms of ecosystem development, OP Stack adopts a top-down operational strategy, successively launching Base, opBNB, and Worldcoin chains through collaborations with well-known companies and projects like Coinbase, BNB Chain, and Worldcoin, quickly gaining user recognition and establishing a foothold in the Layer2 market, thus becoming a model template for the development of L2 Stacks products.

In August of this year, the revenue-sharing cooperation framework agreement jointly announced by Base and Optimism once again made OP Stack a hot topic in industry discussions. According to the proposal, Base can provide Optimism with optional potential revenue methods, paying either 2.5% of the sequencer revenue or 15% of the profits, whichever is higher. In return, Optimism will allocate 2.75% of the total supply of OP tokens (approximately 11.8 million tokens) to the Base chain. This cooperation proposal has deepened the connection between Base and Optimism beyond just the OP Stack technology link; the two will empower each other, with Base benefiting from the backing of Coinbase, which will also feed back into the OP Stack ecosystem.

From the cooperation framework agreement between Base and Optimism, one can glimpse the strategy for OP Stack to realize the Superchain empire roadmap: in the early stages, Optimism helps the early OP Stack chains develop by distributing OP tokens to rapidly expand its ecosystem. Meanwhile, these L2 networks based on OP Stack will share their revenue with Optimism, feeding back into the platform. This positive cycle may propel the rapid expansion of Optimism's flywheel.

Currently, there are dozens of Layer2 networks built on OP Stack, including Coinbase's Base chain, Binance's opBNB, Gitcoin's PGN, Manta Network's Manta Pacific, Zora Network, and more.

*

#### Arbitrum Orbit Focuses on L3 Expansion and Developer Experience

Arbitrum Orbit is an open-source toolkit launched by Arbitrum for building L3 networks, supporting developers in deploying and creating their own application chains.

Developers can build their own Rollup or AnyTrust chains (L3) based on the Arbitrum Orbit toolkit, choosing either Arbitrum One (Arbitrum Rollup chain) or Arbitrum Nova (AnyTrust chain) as the settlement layer.

The main difference between Rollup and AnyTrust is that in Rollup, all transaction data is published on Layer1 (Ethereum mainnet), making it more decentralized, while in AnyTrust, data is managed off-chain, offering better performance suitable for high-frequency trading scenarios.

In simple terms, the settlement layer of networks built using Arbitrum Orbit is not Ethereum but rather the Layer2 networks Arbitrum One or Arbitrum Nova, extending from the Layer2 network, hence referred to as L3 networks, i.e., L1→L2→L3.

If developers want to use Arbitrum Orbit to build Layer2 networks, they need to obtain a license for the relevant code. Currently, Arbitrum still uses a Business Source License (BSL). There are mainly two ways to obtain a license: one is to apply to the initial development team Offchain Labs of the Arbitrum codebase; the other is to propose an application to the Arbitrum DAO, which will then vote on whether to grant the license. The Arbitrum official recommends the second method. It is reported that projects obtaining license authorization may need to pay a code licensing fee to Arbitrum. However, some team members have proposed to "publish the conditions and requirements for using the Arbitrum code license," which has sparked related discussions, and many community users speculate that the Arbitrum license requirements may soon be lifted.

It is important to note that developers using Arbitrum Orbit to build L3 networks do not need any licensing requirements. This is likely because the settlement layer of the L3 network is on Arbitrum One or Arbitrum Nova, allowing Arbitrum to earn some revenue through settlement fees.

Currently, several applications have partnered with Arbitrum Orbit to create their own L3 networks. Among them, the Ethereum scaling service platform AltLayer has supported Arbitrum Orbit, allowing users to launch Arbitrum's L3 application chain in minutes using no-code tools; the Xai network, designed specifically for gaming, announced it will utilize Arbitrum Orbit technology to launch a custom L3 blockchain for the gaming industry; in the same month, the DeFi derivatives trading platform Syndr announced the launch of Syndr Chain based on Arbitrum Orbit, with the testnet already online.

Additionally, Offchain Labs has equipped developers with the smart contract development platform Arbitrum Stylus, which launched its public testnet on August 31. Stylus is a development environment built by Arbitrum for developers, providing tools for building and testing smart contracts on the Arbitrum network, including smart contract development, testing, deployment, and interaction with the Arbitrum network.

Stylus is considered another significant innovation for Arbitrum, as it allows its EVM language to no longer be limited to traditional Solidity, enabling EVM to support more programming languages (such as Rust, C, and C++) for building on-chain applications, which may attract millions of developers into the Web3 ecosystem.

It is reported that there are currently about 20,000 developers using the mainstream EVM programming language Solidity, while there are 3 million Rust developers and 12 million C developers. Additionally, applications based on Rust code development in ecosystems like Solana and Near can also be ported to the Arbitrum ecosystem.

Many believe that the impact of Arbitrum's Stylus will be profound, reaching developers from different programming languages and bringing in more grassroots developer communities, increasing the likelihood of interesting applications emerging from the influx of developers.

In terms of its product mechanism, Arbitrum is also actively exploring innovations. The newly released verification protocol Bold in August allows for permissionless verification on Arbitrum, meaning that the 7-day waiting period for users to withdraw assets from the Arbitrum mainnet to Ethereum will become a thing of the past.

*

#### Differences and Advantages Between OP Stack and Arbitrum Orbit

In terms of product development strategies, the two have clearly different focuses. OP Stack is primarily used to build L2 networks, while Arbitrum Orbit focuses mainly on the L3 market. Users wishing to build L2 networks need to apply for a license and obtain official approval before proceeding.

Additionally, the two differ in their code open licensing methods. Offchain Labs co-founder Steven Goldfeder once explained that OP Stack adopts a "free entry, equipment fee" model, meaning entry to the park is free, but if you want to use the equipment inside, you need to pay an additional fee; Arbitrum Orbit follows a "pay to enter, free to play" model, meaning a fee must be paid before entering the park, allowing free access to all equipment inside, similar to a Disney package.

Specifically, the OP Stack code stack is open-source and free, and using it to build L2 is also free (i.e., free entry). However, if your L2 chain wants to interoperate with OP Mainnet, Base, and more chains in the future, you must obtain a license to be considered part of the Superchain and contribute to the superchain's ecological resources; otherwise, it is just an OP Stack chain. The Superchain license, also known as the "Law of Chains," requires compliance with OP governance. In simple terms, the OP Stack's Superchain ecosystem is a consortium model, and newly built L2 networks can only enjoy Superchain resources, such as cross-chain asset interoperability, by joining this consortium, which may have conditions like staking a certain amount of OP tokens or sharing revenue with OP Stack similar to Base.

However, currently, new chains using OP Stack do not necessarily have to comply with the "Law of Chains" to interoperate with other OP Stack chains (like OP Mainnet, Base, etc.); they can also build their own bridging tools and infrastructure to achieve cross-chain interoperability within the Superchain system, but this undoubtedly increases the workload.

For Arbitrum Orbit, there are two scenarios: L2 and L3. If building L3, it is completely open and requires no permission. However, if building L2, an application must be submitted to the Arbitrum Foundation or DAO organization to obtain a license, and a certain fee must be paid (i.e., pay to enter). However, once a license is obtained, subsequent access to the ecosystem is free, allowing developers complete freedom to manage their chains, choose the software to run, and decide which other chains to interoperate with.

The different charging methods also determine the differences in revenue models between Arbitrum Orbit and OP Stack. Arbitrum Orbit attempts to profit through its proprietary development environment codebase, while OP Stack seems to aim to profit through its shared sequencer infrastructure.

ZK Camp: Supporting L2 and L3 with ZK Stack, Unreleased Starknet Stack, and Latecomer Polygon CDK

*

#### Supporting the Construction of ZK Camp L2 and L3 with ZK Stack

ZK Stack is an open-source, composable, and free modular framework developed by zkSync, enabling developers to build customized ZK-supported L2 and L3 networks (also referred to as Hyperchains) based on the zkSync Era mainnet code. Any developer can freely customize Hyperchains according to their specific needs.

When developers use ZK Stack to build L2 networks, like OP Stack, security is fully inherited from Ethereum. However, according to Matter Labs' official introduction, developers enjoy complete autonomy in choosing the economic token model for the sequencer and data availability mode (currently, OP Stack chains only support using ETH as the on-chain gas fee payment token), allowing developers to decide freely. Developers have full rights to the ZK Stack code, enabling them to customize and shape various aspects of the chain.

In terms of cross-chain communication, zkSync introduces the concept of Hyperbridges, which enables cross-chain communication between L2 and L3 Superchains, similar to how hyperlinks in the internet world can instruct a browser to automatically open a specific webpage.

Hyperbridges consist of a series of smart contracts deployed on L1 that can verify the Merkle tree proofs of transactions occurring on other chains. Original assets will be locked in shared bridge contracts on L1, and when there is a need for cross-chain asset transfer, the shared bridge contract will manage and allocate them uniformly, meaning the liquidity of the entire Superchain ecosystem is unified, and assets, data, and messages can be managed uniformly by the shared bridge contracts deployed on L1.

Additionally, in the ZK Stack Superchain ecosystem, each Superchain supports seamless integration, with unified liquidity. In this ecosystem, users can treat multiple L2 and L3 networks as a whole "Superchain," allowing them to move assets between different chains without switching RPC networks or undergoing cross-chain steps, capturing more revenue opportunities.

*

#### Unreleased Starknet Stack

Starknet announced the launch of the Starknet Foundry, an open-source toolkit for building application chains (Appchains), at the Paris EthCC conference in July this year. Based on the local Cairo test runner and Blockifie written in Rust, Starknet Foundry supports any application to deploy its Starknet application chain in a permissionless manner.

According to the official GitHub page, the Starknet Foundry product is still under development and has not yet been officially released to the public.

*

#### Latecomer Polygon CDK

On August 31, Polygon officially announced the launch of the Polygon CDK (Polygon Chain Development Kit), a standardized code library for development that anyone can use to build ZK technology-driven L2 networks while inheriting Ethereum's security.

Polygon CDK is an upgraded version of Polygon Supernet (an open-source toolkit for building L1). Now, leveraging Polygon's ZK technology, developers can easily customize and deploy their ZK-based L2 application chains using Polygon CDK. Additionally, Polygon supports upgrading the current architecture of all existing Supernet networks to Polygon ZK technology. In other words, Polygon CDK helps developers launch new ZK-based L2 chains or seamlessly transition existing L1 chains to L2.

Moreover, Polygon CDK aims to provide developers with a flexible and customizable modular environment, allowing them to choose CDK modular components to customize their chains according to their needs. For example, developers can select the VM mode (zkEVM or other ZK-supported execution environments like MidenVM), sequencer type (centralized or decentralized sequencer mode), transaction data storage (Rollup on-chain or validium off-chain mode), data availability solutions, and set the gas token to their native token.

In terms of cross-chain communication, the chains created through CDK are interconnected, belonging to the same layer as Polygon PoS and Polygon zkEVM, enabling resource interoperability and sharing.

Within less than half a month of its launch, two chains have chosen to adopt the Polygon CDK solution to build L2 networks. On September 13, the smart contract platform Astar Network in the Polkadot ecosystem announced a partnership with Polygon Labs to launch the Ethereum Layer2 network Astar zkEVM. According to the latest statements from Astar Network's founder, the zkEVM testnet will launch this month; on September 19, the Cosmos ecosystem EVM public chain Canto announced it would build ZK L2 based on Polygon CDK; on October 5, the zkEVM network launched by the gaming public chain Immutable, which previously collaborated with Polygon, announced it would undergo an upgrade and reconstruction, with the zkEVM mainnet set to launch between December and January next year. Additionally, Polygon Labs has proposed to Celo to use the Polygon CDK suite to migrate to L2 and suggested that ApeCoin DAO use Polygon CDK to develop the L2 network "ApeChain."

What Considerations Do Developers Have When Choosing L2 Stacks Solutions?

Faced with the five mainstream L2 Stacks solutions—OP Stack, Arbitrum Orbit, Polygon CDK, ZK Stack, and Starknet Stack—what criteria do developers use to choose a solution?

Currently, developers primarily evaluate an L2 Stacks solution based on factors such as "compatibility with Ethereum, the number of deployed projects, cross-chain communication costs, and the entry barrier (also known as code openness)."

Among these, compatibility with Ethereum will determine the interoperability and development difficulty between the L2 network built on L2 Stacks and Ethereum. After all, most projects choose to build L2 networks to inherit Ethereum's security and hope to share in Ethereum's massive traffic. Currently, in terms of product underlying compatibility, OP Stack and Arbitrum Orbit undoubtedly have better compatibility with Ethereum than ZK camp's Stacks products, as zero-knowledge proofs have always been a challenge for compatibility with the Ethereum Virtual Machine. Although the zkEVM mainnet has been launched, there are still shortcomings in compatibility with Ethereum.

The entry barrier for using L2 Stacks will directly affect the complexity of adoption by developers and also represents the code openness of the toolkit.

Currently, the entry barriers mainly include costs and development expenses. The former primarily refers to whether using an L2 Stack requires paying certain fees to the official. For instance, OP Stack adopts a "free entry, equipment fee" model, initially offering free support for developers to build L2 networks. However, if these networks want to enjoy resource sharing with Base and OP Mainnet, they need to join relevant alliance organizations, which have certain entry barriers and reportedly charge fees. Arbitrum Orbit adopts a "pay to enter, free to play" strategy, requiring official licensing and payment of fees to build L2, but no further fees are needed afterward.

For early-stage developers, the free entry barrier of OP Stack is lower; the latter development costs mainly refer to the technical and capability requirements for developers. For example, ZK camp's L2 Stack may require developers to understand zero-knowledge proof technology. The Starknet Stacks open-source toolkit is based on a local Cairo test runner and Blockifie written in Rust, and developers' familiarity with related knowledge will affect the project's development progress.

The feasibility of landing and the number of deployed projects represent the maturity of the L2 Stacks solution. Currently, OP Stack and Polygon CDK have more deployed projects on L2 networks, while Arbitrum Orbit has more applications deployed on L3.

The ease of inter-chain communication within the Superchain ecosystem and cross-chain communication with external heterogeneous chains will determine the future interoperability of the chains. However, currently, asset cross-chain transfers between L2 networks within the Superchain ecosystem have not been fully established. For instance, the official cross-chain bridge of OP Stack's Base only supports transfers with Ethereum and does not support cross-chain transfers of assets from the OP Mainnet chain. As for cross-chain communication between the Superchain and external chains like Binance Chain and Solana, there is currently no comprehensive solution apart from third-party cross-chain bridges.

Additionally, we can draw some conclusions from certain projects when choosing to adopt L2 Stack. The privacy network Manta, which has launched the EVM execution layer Manta Pacific based on OP Stack, went live with its mainnet on September 13. According to browser data, the mainnet has recorded 430,000 transactions and over 40,000 wallet addresses, with data consistently on the rise.

Originally named Manta Network, Manta is a ZK privacy Layer1 public chain in the Polkadot ecosystem. In July this year, it announced the launch of the Layer2 network EVM native modular execution layer Manta Pacific, designed specifically for zero-knowledge proof (ZK) applications based on OP Stack, and renamed its Layer1 network Manta to Manta Atlantic. The Manta ecosystem will consist of Manta Atlantic and Manta Pacific.

Regarding the choice of which L2 Stack to adopt, a relevant person from Manta stated that since the application circuits are designed based on Ethereum, the main considerations when choosing an L2 Stack solution are compatibility with Ethereum and potentially the ecosystem and stability of the technology stack.

Currently, OP Stack and Polygon CDK have the most deployments and are relatively mature solutions. Among them, OP Stack has projects like Base and opBNB built on it, and OP Stack has also established reliable integration solutions with Celestia. Having a cheaper DA like Celestia is essential for Manta's ecosystem projects to operate more affordably. We are also exploring some more innovative architectures beyond OP Stack, and besides using Celestia, there are also some innovations in scaling. Polygon CDK has only emerged in the last two months.

Although Arbitrum has solid technology and a large user base with Arbitrum One, the process for issuing L2 is somewhat complex, requiring an application to be submitted to the official for approval before deployment. Additionally, the existing zkEVM solutions are not yet very mature; for instance, while zkSync's ZK Stack has been launched, it has not yet seen representative deployed projects, and its compatibility with Ethereum still needs improvement. Currently, developers need to use two sets of code for ZK Stack and the Ethereum mainnet. Starknet represents a relatively independent implementation approach and ecosystem, which has not yet officially launched.

From this perspective, OP Stack is currently the most popular L2 Stack, followed by Polygon CDK and Arbitrum. However, these L2 Stack solutions are not without drawbacks; since the settlement layer and security are guaranteed by Ethereum, any issues with Ethereum's operation will also affect the operation of these L2 Stack networks.

As for how to choose a suitable L2 Stack, developers can also leverage some third-party RaaS service platforms, which can help evaluate whether certain solutions fit their needs and integrate the advantages of various L2 Stacks for customization. However, third parties often charge high fees. It is reported that some third-party service providers may charge L2 networks service fees exceeding 30%.