No Free Lunch: Reflections on Arbitrum and Optimism

Key Takeaways

• Base announced a transition away from Optimism's OP Stack toward a proprietary unified architecture, sending a strong shockwave through the market and hitting $OP hard.

• Optimism fully opens its code under an MIT license and operates a revenue-sharing model for chains that join the Superchain. Arbitrum adopts a "Community Source" model that requires chains built on Orbit to contribute 10% of protocol revenue if they settle outside the Arbitrum ecosystem.

• The debate over open-source monetization in blockchain infrastructure is an extension of the same problem that has played out repeatedly in traditional software with Linux, MySQL, MongoDB, WordPress, and others. Yet the introduction of tokens as a variable creates an additional layer of stakeholder dynamics.

• It is difficult to declare either side definitively right. What matters is understanding the trade-offs each model entails with clear eyes, and thinking collectively as an ecosystem about the long-term sustainability of L2 infrastructure.

1. Base's Departure and Cracks in the Superchain

On February 18, Coinbase's Ethereum L2 network Base announced that it would sever its dependency on Optimism's OP Stack and transition to a proprietary unified codebase. The core idea is to consolidate key components, including the sequencer, into a single repository while reducing reliance on external dependencies such as Optimism, Flashbots, and Paradigm. In its official blog post, the Base engineering team stated that this transition would increase the annual hard fork cadence from three to six, effectively doubling the pace of upgrades.

The market's reaction was immediate: $OP fell more than 20% within 24 hours. Given that the largest chain in Optimism's Superchain ecosystem had just declared independence, this was hardly surprising.

Source: @sgoldfed

Around the same time, Steven Goldfeder, co-founder of Arbitrum and CEO of Offchain Labs, posted on X reminding people that his team had deliberately chosen a different path years ago. The gist of his message was that despite pressure to release Arbitrum's code as fully open source, the team had stuck with what they call the "Community Source" model.

Under this model, the code itself is publicly available, but any chain built on Arbitrum's Orbit stack is required to contribute a fixed percentage of its protocol revenue to the Arbitrum DAO. Goldfeder added a pointed warning: "If a stack allows extraction without contribution, eventually that's exactly what will happen."

Base's departure is not merely a technical migration. It is an event that has brought a fundamental question to the surface: what kind of economic structure should blockchain infrastructure be built upon? In this piece, we will examine the economic frameworks adopted by Optimism and Arbitrum, explore how they differ, and discuss where the industry should be headed.

2. Two Models

Optimism and Arbitrum handle their software in fundamentally different ways. Both are leading projects in Ethereum's L2 scaling landscape, but they diverge sharply in how they approach the economic sustainability of their ecosystems.

2.1 Optimism: Openness and Network Effects

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

Revenue sharing only kicks in when a chain joins the Superchain, Optimism's official ecosystem. Superchain members contribute the greater of 2.5% of chain revenue or 15% of on-chain net income (fee revenue minus L1 gas costs) to the Optimism Collective. In return, they gain access to the Superchain's shared governance, shared security, interoperability, and brand.

The logic underpinning this approach is straightforward. If countless L2 chains are built on the OP Stack, those chains form an interoperable network, and through network effects, the value of the OP token and the Optimism ecosystem as a whole rises. In practice, this strategy has delivered substantial results. Major projects including Coinbase's Base, Sony's Soneium, Worldcoin's World Chain, and Uniswap's Unichain have all adopted the OP Stack.

There are reasons beyond the licensing model that explain why large enterprises have favored the OP Stack. Alongside the freedom afforded by the MIT license, the OP Stack's modular architecture is a core competitive advantage. Because the execution layer, consensus layer, and data availability layer can each be swapped out independently, projects like Mantle and Celo have been able to adopt zero-knowledge proof modules such as OP Succinct and customize at will. The ability to take the code without external permission and freely replace internal components was particularly attractive from a corporate sovereignty standpoint.

However, the structural vulnerability of this model is equally clear: 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 more profitable a chain becomes, the more economically rational it is to operate independently. Base's departure is a textbook illustration of this dynamic.

2.2 Arbitrum: Enforced Alignment

Arbitrum takes a more complex approach. For L3 chains built on Arbitrum Orbit that settle on Arbitrum One or Nova, there is no revenue-sharing obligation. However, under the Arbitrum Expansion Program (AEP), chains that settle on a network other than Arbitrum One or Nova, whether L2 or L3, are required to contribute 10% of net protocol revenue to Arbitrum. Of that 10%, 8% goes to the Arbitrum DAO treasury and 2% to the Arbitrum Developer Guild.

In other words, chains that remain within the Arbitrum ecosystem are given freedom, while chains that take Arbitrum's technology and deploy it in an external ecosystem are required to contribute. It is a dual structure.

In the early days, building an L2 on Arbitrum Orbit that settles directly on Ethereum required approval through Arbitrum DAO governance vote. When the AEP was introduced in January 2024, this process shifted to a self-service model. Still, the early "permissioned" process and the emphasis on encouraging L3s likely acted as a barrier for large enterprises seeking sovereign L2 chains. For companies wanting an L2 that connects directly to Ethereum, an L3 structure sitting on top of Arbitrum One carries additional business risk in the form of governance and technical dependency.

Goldfeder's decision to label this model "Community Source" is intentional. It positions itself as a third path between traditional open source and proprietary licensing. Code transparency is preserved, but using that code commercially outside the Arbitrum ecosystem comes with a mandatory contribution to the ecosystem.

The strength of this model lies in aligning the economic interests of ecosystem participants. There is a tangible exit cost for chains settling externally, and a sustainable revenue stream is secured. The Arbitrum DAO has reportedly accumulated approximately 20,000 ETH in revenue, and Robinhood's recent announcement that it would build its own L2 chain on Orbit has further validated the model's potential for institutional adoption. Robinhood's chain testnet recorded 4 million transactions in its first week, demonstrating that Arbitrum's technical maturity and regulatory-friendly customization capabilities provide meaningful value to certain types of institutional clients.

2.3 The Trade-offs of Each Model

The two models optimize for different values. Optimism's model maximized the speed of initial enterprise adoption through the unconditional openness of its MIT license, its modular architecture, and the powerful proof of concept that Base represented. An environment where code can be taken without permission, components can be swapped freely, and a proven reference case already exists offered business decision-makers the lowest possible entry threshold.

Arbitrum's model, on the other hand, emphasizes long-term ecosystem sustainability. In addition to technical excellence, its economic alignment mechanism requiring revenue contributions from external users secures a stable funding base for infrastructure maintenance. Initial adoption speed may be somewhat slower, but for projects built using capabilities unique to the Arbitrum stack, such as Arbitrum Stylus, the exit costs can be significant.

That said, the difference between the two models is not as extreme as it is often portrayed. Arbitrum also offers a free and permissionless license within its ecosystem, and Optimism also requires revenue sharing for Superchain membership. Both sit on a spectrum between "fully open" and "fully enforced," differing in degree and scope rather than in kind.

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

3. Lessons from Open-Source History

This tension is not unique to blockchain. Open-source software monetization models have gone through remarkably similar debates over the past several 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 penetrated nearly every domain of computing: servers, cloud, embedded systems, Android, and more.

Yet the most commercially successful business built on top of this ecosystem, Red Hat, did not monetize the code itself. It monetized services built on top of that code. Red Hat sold enterprise customers technical support, security patches, and stability guarantees, and was acquired by IBM for $34 billion in 2019. The code is free, but professional operational support comes at a cost. This logic bears a striking resemblance to Optimism's recently launched OP Enterprise.

3.2 MySQL and MongoDB

MySQL introduced a dual licensing model: an open-source version under the GPL, alongside a separate commercial license sold to enterprises that wanted to use MySQL for commercial purposes. The code is visible and free for non-commercial use, but generating revenue from it requires payment. The concept is similar to Arbitrum's Community Source model.

MySQL succeeded with this approach, but it was not without side effects. When Oracle acquired Sun Microsystems in 2010 and with it ownership of MySQL, concerns about MySQL's future led its original creator Monty Widenius and community developers to create MariaDB, a fork. While it was the change in ownership structure rather than licensing policy that served as the direct catalyst, the possibility of forking is an ever-present risk in open-source software. The parallel to Optimism's current situation is hard to miss.

MongoDB offers an even more direct example. In 2018, MongoDB adopted the Server Side Public License (SSPL). The motivation was to address a growing problem: large cloud providers like AWS and Google Cloud were taking MongoDB's code, offering it as a managed service, and paying nothing back to MongoDB. Actors who extract value from open code without contributing anything in return: this is a pattern that recurs throughout open-source history.

3.3 WordPress

WordPress is fully open source under the GPL license and powers roughly 40% of all websites worldwide. Automattic, the company behind WordPress, generates revenue through the WordPress.com hosting service and various plugins, but it charges nothing for the use of WordPress core itself. The platform is fully open, and the logic is that the growth of the ecosystem itself elevates the platform's value. This is structurally similar to Optimism's Superchain vision.

The WordPress model has clearly succeeded. But the free-rider problem was never fundamentally resolved. In recent years, a dispute erupted between WordPress creator Matt Mullenweg and the major hosting company WP Engine. Mullenweg publicly criticized WP Engine for generating enormous revenue from the WordPress ecosystem while contributing insufficiently in return. The paradox where the biggest beneficiary of an open ecosystem contributes the least: this is the exact same dynamic that played out between Optimism and Base.

4. Why Crypto Is Different

These debates have already been well-rehearsed in traditional software. So why does the problem become particularly acute in blockchain infrastructure?

4.1 Tokens as an Amplifier

In traditional open-source projects, value is relatively diffuse. When Linux succeeds, no specific asset's price rises or falls in direct response. But in blockchain ecosystems, tokens exist, and tokens reflect the incentives and political dynamics of ecosystem participants in real time through price.

In conventional open-source software, free-riding leads to a shortage of development resources. That is serious, but the consequences unfold gradually. In blockchain, the departure of a major participant triggers an immediate and highly visible outcome: a token price crash. The fact that $OP dropped more than 20% after Base's announcement illustrates this clearly. Tokens serve as both a barometer of ecosystem health and a mechanism that amplifies crises.

4.2 The Responsibility of Financial Infrastructure

L2 chains are not just software. They are financial infrastructure. Billions of dollars in assets are managed on top of them, and maintaining their stability and security requires enormous ongoing costs. In successful open-source projects, maintenance costs are often covered by corporate sponsorship or foundation support, but most L2 chains today are already stretched thin just running their own ecosystems. Without external contributions in the form of sequencer fee sharing, it becomes difficult to secure the resources needed for infrastructure development and maintenance.

4.3 Ideological Tension

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

5. Closing Thoughts: There Is No Free Infrastructure

It is true that Base's departure dealt a blow to Optimism, but it would be premature to read it as a failure of the Superchain model itself.

Source: @Optimism

For one, Optimism has not been sitting idle. On January 29, 2026, Optimism officially launched OP Enterprise, an enterprise service targeting fintechs and financial institutions that enables deployment of production-grade chains within 8 to 12 weeks. While the vanilla OP Stack is MIT-licensed and can always be transitioned to a self-managed model, Optimism's assessment is that for most teams who are not blockchain infrastructure specialists, partnering with OP Enterprise is the more rational choice.

Nor is Base severing ties with the OP Stack overnight. Base itself has stated that it will remain an OP Enterprise mission-critical support customer during the transition period and plans to maintain compatibility with OP Stack specifications throughout. The separation is technical, not relational. That is the official position from both sides. On the other hand, Arbitrum's Community Source model also has gaps between its ideals and reality.

The roughly 19,400 ETH in net fee revenue accumulated in the Arbitrum DAO treasury has, in fact, come almost entirely from sequencer fees and Timeboost MEV auctions on Arbitrum One and Nova themselves. Fee-sharing revenue contributed by ecosystem chains through the AEP has not been publicly confirmed at any meaningful scale. There are structural reasons for this. The AEP itself was only introduced in January 2024, the majority of existing Orbit chains are L3s built on Arbitrum One and thus exempt from revenue-sharing obligations, and even the most prominent AEP-eligible independent L2, the Robinhood chain, remains in testnet.

For Arbitrum's Community Source model to carry real weight as a "sustainable revenue structure," the ecosystem will need to wait until large L2s like Robinhood reach mainnet and AEP fee-sharing revenue actually begins flowing in. The requirement to hand over 10% of protocol revenue to an external DAO is not an easy decision for a large enterprise. The fact that an institution like Robinhood chose Orbit regardless speaks to value propositions on other dimensions, namely customization potential and technical maturity. But the economic justification of the model remains unproven. The gap between theoretical design and actual revenue flows is a challenge Arbitrum still needs to address.

The two models offered by Arbitrum and Optimism are, in the end, different answers to the same question: how do you ensure the sustainability of public infrastructure?

What matters is not which model is right, but understanding the trade-offs each model carries. Optimism's open model enabled rapid ecosystem expansion but carried the inherent risk that its biggest beneficiary might leave. Arbitrum's enforced contribution model creates a sustainable revenue structure but raises the entry threshold for initial adoption.

Whether we are talking about Optimism or Arbitrum, OP Labs, Sunnyside Labs, and Offchain Labs all employ world-class research talent dedicated to scaling Ethereum while preserving decentralization. The technical progress of L2 scaling is impossible without their continued development investment, and the resources to fund that work have to come from somewhere.

There is no free infrastructure. What we should do as a community is not pledge blind loyalty or harbor knee-jerk resentment toward any particular model, but rather start an honest conversation about who bears the cost of this infrastructure. Base's departure can be the starting point for that conversation.