Original title : [Issue] No Free Lunch: Reflections on Arbitrum and Optimism

Original author: Four Pillars

Original translation by: Ken, ChainCatcher

Key Summary

Base's announcement of its transition from Optimism's OP stack to a proprietary unified architecture has sent shockwaves through the market and severely impacted the price of $OP.
Optimism is fully open-source under the MIT license and implements a revenue-sharing model for chains that join the "Hyperchain". Arbitrum adopts a "community source code" model, requiring chains built on Orbit to contribute 10% of the protocol revenue if they settle outside the Arbitrum ecosystem.
The debate over open-source monetization in blockchain infrastructure is an extension of recurring issues in traditional software fields (such as Linux, MySQL, MongoDB, WordPress, etc.). However, the introduction of tokens as variables adds another layer of dynamic relationships among stakeholders.
It's difficult to say which side is absolutely right. What's important is to have a clear understanding of the trade-offs involved in each model and to think collectively as an ecosystem about the long-term sustainability of L2 infrastructure.

1. The departure of Base and the cracks in the superchain

On February 18th, Coinbase's Ethereum L2 network, Base, announced it would sever its reliance on the Optimism operating stack and transition to a proprietary unified codebase. The core idea is to integrate key components, including the sequencer, into a single repository, while reducing external dependencies on Optimism, Flashbots, and Paradigm. The Base engineering team stated in their official blog that this shift will increase the frequency of hard forks per year from three to six, effectively accelerating upgrades.

The market reacted swiftly: $OP fell more than 20% in 24 hours. This is not surprising, considering that the largest chain in the Optimism superchain ecosystem has just announced its independence.

Around the same time, Steven Goldfeder, co-founder of Arbitrum and CEO of Offchain Labs, posted on the X platform, reminding everyone that his team deliberately chose a different path a few years ago. His core point was that despite pressure to release the Arbitrum code as fully open source, the team insisted on what they called the "community source code" model.

In this model, the code itself is public, but any chain built on the Arbitrum Orbit stack is required to contribute a fixed percentage of protocol revenue to the Arbitrum decentralized autonomous organization. Goldfeder issued a stark warning: "This is what will happen if a stack allows revenue to be claimed without contributing."

Base's departure is more than just a technological migration. This event brings a fundamental question to the forefront: on what economic structure should blockchain infrastructure be built? This article will examine the economic frameworks adopted by Optimism and Arbitrum, explore their differences, and discuss the future direction of the industry.

2. Two modes

Optimism and Arbitrum approach software in drastically different ways. Both are leading projects in the Ethereum L2 scaling space, but they diverge significantly on their methods for achieving economic sustainability within their ecosystems.

2.1 Optimism: Openness and Network Effects

Optimism's OP stack is completely open source under the MIT license. Anyone can obtain the code, modify it freely, and build their own L2 chain. There are no royalties or revenue-sharing obligations.

Revenue sharing is only activated when a chain joins Optimism's official ecosystem, the "Hyperchain." Members must contribute either 2.5% of their chain's revenue or 15% of their on-chain net revenue (fee revenue minus Layer 1 network gas costs), whichever is higher, to the Optimism Collective. In return, they gain access to the Hyperchain's shared governance, shared security, interoperability, and brand resources.

The logic behind this approach is simple. If numerous L2 chains are built on the OP stack, these chains will form an interoperable network, and through network effects, the value of the OP token and the entire Optimism ecosystem will increase. In fact, this strategy has already yielded significant results. Major projects such as Coinbase's Base, Sony's Soneium, Worldcoin's World Chain, and Uniswap's Unichain have all adopted the OP stack.

Large enterprises favor the OP stack for reasons beyond just licensing models. Beyond the freedom offered by the MIT license, the modular architecture of the OP stack is a core competitive advantage. Because the execution layer, consensus layer, and data availability layer can be independently replaced, projects like Mantle and Celo can adopt and freely customize zero-knowledge proof modules such as OP Succinct. For enterprise sovereignty, the ability to obtain code without external licenses and freely replace internal components is extremely attractive.

However, the structural weaknesses of this model are equally apparent: low barriers to entry also mean low barriers to exit. Chains using the OP stack have limited economic obligations to the Optimism ecosystem, and the higher the chain's profits, the more economically rational independent operation becomes. Base's departure is a textbook example of this dynamic.

2.2 Arbitrum: Forced Coordination

Arbitrum takes a more complex approach. For L3 chains built on Arbitrum Orbit and settled on Arbitrum One or Nova, there is no revenue-sharing obligation. However, under Arbitrum's scaling plan, chains that settle on networks other than Arbitrum One or Nova (whether Layer 2 or Layer 3) are required to contribute 10% of their net protocol revenue to Arbitrum. Of this 10%, 8% goes to the Arbitrum Decentralized Autonomous Organization Treasury, and 2% goes to the Arbitrum Developer Association.

In other words, chains that remain within the Arbitrum ecosystem enjoy freedom, while chains that utilize Arbitrum technology and are deployed in external ecosystems must contribute. This is a dual structure.

Initially, building the Arbitrum Orbit L2, which settles directly on Ethereum, required approval through a governance vote within the Arbitrum decentralized autonomous organization. This process transitioned to a self-service model when the Arbitrum expansion plan launched in January 2024. Nevertheless, the early "permissioned" process and the emphasis on encouraging L3 may have been obstacles for large enterprises seeking a sovereign L2 chain. For companies looking to connect directly to Ethereum, an L3 architecture built on Arbitrum One presents additional business risks in terms of governance and technological dependence.

Goldfeder's decision to call this model "community source" was intentional. It positions itself as a third path between traditional open source and proprietary licenses. Code transparency is preserved, but commercial use outside the Arbitrum ecosystem requires contributions to the ecosystem.

The advantage of this model lies in aligning the economic interests of ecosystem participants. For chains settling externally, there are tangible exit costs, thus ensuring a sustainable revenue stream. The Arbitrum decentralized autonomous organization has reportedly accumulated approximately 20,000 ETH in revenue, and Robinhood's recent announcement of building its own L2 chain on Orbit further validates the model's potential for institutional adoption. The Robinhood Chain testnet recorded 4 million transactions in its first week, demonstrating Arbitrum's technological maturity and regulatory-friendly customization capabilities, providing meaningful value to specific types of institutional clients.

2.3 Trade-offs among the various models

The two models are optimized for different values. The Optimism model maximizes the speed of early enterprise adoption through the unconditional openness of the MIT license, modular architecture, and the strong proof-of-concept represented by Base. An environment that allows for licensed access to code, free replacement of components, and mature reference cases provides business decision-makers with the lowest possible barrier to entry.

On the other hand, Arbitrum's model emphasizes the sustainability of its long-term ecosystem. In addition to its superior technology, its economic coordination mechanism requires external users to contribute revenue, ensuring a stable financial foundation for infrastructure maintenance. Initial adoption may be slightly slower, but for projects built using the unique features of the Arbitrum stack, such as Arbitrum Stylus, exit costs can be quite high.

That said, the differences between these two models are not as extreme as they are often described. Arbitrum also offers free and permissionless licenses within its ecosystem, and Optimism requires superchain members to share revenue. Both lie on the spectrum between “completely open” and “completely mandatory,” differing in degree and scope rather than in essence.

Ultimately, this difference is a blockchain version of the classic trade-off between growth rate and sustainability.

3. Lessons from the history of open source

This tension is not unique to blockchain. Open-source software monetization models have experienced remarkably similar debates over the past few decades.

3.1 Linux and Red Hat

Linux is the most successful open-source project in history. The Linux kernel is fully open under the GPL license and has permeated almost every area of computing: servers, cloud, embedded systems, Android, and more.

However, Red Hat, the most successful commercial enterprise built on this ecosystem, does not profit from the code itself. It profits from the services built on top of that code. Red Hat sells technical support, security patches, and stability guarantees to enterprise customers and was acquired by IBM for $34 billion in 2019. The code is free, but professional operational support is charged for. This logic bears a striking resemblance to Optimism's recently launched OP Enterprise.

3.2 MySQL and MongoDB

MySQL has adopted a dual-licensing model: an open-source version under the GPL license, and a separate commercial license sold to companies that wish to use MySQL for commercial purposes. The code is visible and free for non-commercial use, but revenue generated from it requires payment. This concept is similar to Arbitrum's community-source model.

MySQL succeeded through this approach, but it wasn't without its side effects. When Oracle acquired Sun Microsystems in 2010 and subsequently gained ownership of MySQL, concerns about its future led its original creator, Monty Widenius, and community developers to create the fork MariaDB. While the immediate catalyst was a change in ownership structure rather than licensing policy, the possibility of forking is a pervasive risk in open-source software. The similarity to Optimism's current predicament is readily apparent.

MongoDB provides a more direct example. In 2018, MongoDB adopted a server-side public license. The motivation was to address a growing problem: cloud service giants like Amazon Web Services and Google Cloud use MongoDB's code, offering it as a managed service, without paying MongoDB any fees. Actors who demand value from open-source code without giving anything in return: this is a recurring pattern throughout the history of open source.

3.3 WordPress

WordPress, fully open source under the GPL license, powers approximately 40% of websites worldwide. Automattic, the company behind WordPress, generates revenue through its WordPress.com hosting service and various plugins, but charges no fees for using the WordPress core itself. The platform is completely open, and the logic is that the growth of the ecosystem itself will increase the platform's value. This is structurally similar to Optimism's hyperchain vision.

The WordPress model has clearly been successful. But the "free-rider" problem has never been fundamentally solved. In recent years, a dispute has erupted between WordPress founder Matt Mullenweg and the main hosting company, WP Engine. Mullenweg has publicly criticized WP Engine for taking huge profits from the WordPress ecosystem but contributing too little in return. This paradox of the biggest beneficiary of an open ecosystem contributing the least is exactly the same dynamic that is happening between Optimism and Base.

4. Why are different in the crypto field?

These debates are commonplace in traditional software. So why does this issue become particularly acute in blockchain infrastructure?

4.1 Tokens as Amplifiers

In traditional open-source projects, value is relatively dispersed. When Linux succeeds, the price of no specific asset directly rises or falls as a result. However, in the blockchain ecosystem, tokens exist, and their prices reflect the incentives and political dynamics of ecosystem participants in real time.

In traditional open-source software, while the problem of free-riding leading to a shortage of development resources is severe, the consequences are gradual. In blockchain, however, the departure of major players triggers immediate and highly visible results: a sharp drop in token prices. The more than 20% drop in $OP after Base's announcement clearly illustrates this point. Tokens are both a barometer of the ecosystem's health and a mechanism that amplifies crises.

4.2 Responsibilities of Financial Infrastructure

Level 2 blockchains are more than just software. They are financial infrastructure. Billions of dollars in assets are managed on these chains, and maintaining their stability and security requires enormous and ongoing costs. In successful open-source projects, maintenance costs are often covered by corporate sponsorship or foundation support, but most Level 2 blockchains today struggle just to keep their own ecosystem running. Without external contributions in the form of sequencer fee sharing, it is difficult to secure the resources needed for infrastructure development and maintenance.

4.3 Ideological Tension

The crypto community has a strong ideological tradition of "code should be free." Decentralization and freedom are core values closely intertwined with the industry's identity. In this context, Arbitrum's fee-sharing model may provoke resistance from some community members, while Optimism's open model is ideologically appealing but faces real challenges regarding economic sustainability.

5. Conclusion: There is no such thing as free infrastructure.

While Base's departure was a blow to Optimism, it would be premature to conclude that the hyperchain model itself has failed.

First, Optimism didn't sit idly by. On January 29, 2026, Optimism officially launched OP Enterprise, an enterprise-grade service for fintech companies and financial institutions, supporting the deployment of production-grade blockchains within 8 to 12 weeks. While the original OP stack is licensed by MIT and can always be converted to a self-governed model, Optimism's assessment is that partnering with OP Enterprise is a more rational choice for most teams that are not blockchain infrastructure experts.

Base will not sever ties with the OP stack overnight. Base itself has stated that it will remain a core support service customer for OP Enterprise during the transition and plans to maintain compatibility with the OP stack specifications throughout the process. This separation is technical, not relational. This is the official position of both parties. On the other hand, there is also a gap between the ideal and reality of Arbitrum's community-driven open-source model.

In reality, the approximately 19,400 ETH of net fee revenue accumulated in the Arbitrum decentralized autonomous organization's treasury comes almost entirely from sequencer fees from Arbitrum One and Nova itself, and from the Timeboost maximum extractable value auction. Fee-sharing revenue contributed by ecosystem chains through the Arbitrum scaling initiative has not yet received any meaningful public confirmation. There are structural reasons for this. The Arbitrum scaling initiative itself only launches in January 2024, most existing Orbit chains are L3 built on top of Arbitrum One, thus exempt from revenue-sharing obligations, and even the most prominent independent L2 chain eligible for the Arbitrum scaling initiative—Robinhood—is still in the testnet phase.

For Arbitrum's community-driven open-source model to truly hold weight as a "sustainable revenue structure," the ecosystem needs to wait for large L2 servers like Robinhood to launch their mainnets, and for the revenue from Arbitrum's scaling program fee sharing to actually begin flowing in. Requiring 10% of the protocol's revenue to be handed over to external decentralized autonomous organizations (DAOs) is no easy feat for large enterprises. The fact that institutions like Robinhood still choose Orbit speaks to its value proposition in other dimensions, namely customization potential and technological maturity. However, the economic viability of this model remains unproven. The gap between theoretical design and actual cash flow is a challenge that Arbitrum still needs to address.

The two models offered by Arbitrum and Optimism are ultimately different answers to the same question: how to ensure the sustainability of public infrastructure?

The important thing is not which model is right, but understanding the trade-offs each model brings. Optimism's open model enables rapid ecosystem expansion, but also carries the inherent risk that its biggest beneficiaries may leave. Arbitrum's mandatory contribution model establishes a sustainable revenue structure, but raises the barrier to initial adoption.

Whether discussing Optimism or Arbitrum, OP Labs, Sunnyside Labs, and Offchain Labs employ world-class research talent dedicated to scaling Ethereum while maintaining decentralization. Technological advancements in L2 scaling would be impossible without their continued development investment, and the resources funding this work must come from somewhere.

There is no such thing as free infrastructure. As a community, what we need to do is not blind loyalty or subconscious resentment, but to start an honest dialogue to discuss who should bear the cost of this infrastructure. Base's departure can be the starting point for this dialogue.

5月Base网络实现链上利润698万美元，领先各L2网络

Popular Articles