Rollup Stacks vs L2s: The New Layer of Competition
Tracking individual L2s means watching the wrong race. The stack beneath them – OP Stack, Arbitrum Orbit, ZK Stack – is where the real competition plays out.
Key takeaways
- A rollup stack is the shared software framework that defines the security model, interoperability rules, and upgrade path for any chain built on it.
- An individual L2 is a single network that inherits those rules while making its own decisions about use case, governance, and user experience.
- Multiple independent L2s can run on the same stack, meaning a stack's success benefits all chains on it, and its vulnerabilities expose them all equally.
- After EIP-4844 compressed fees across all networks, the real competitive differentiator shifted to ecosystem composability – a stack-level property.
A rollup stack (like OP Stack, Arbitrum Orbit, or ZK Stack) is the shared software framework used to deploy and run multiple L2 chains. An individual L2 is a single network built on top of that framework. The stack defines the rules, including security model, interoperability, and upgrade path, while the L2 plays by them.
Most people tracking Ethereum scaling focus on individual chains, comparing TVL, fees, and user counts. But the more consequential competition is happening one layer below.
What Sets a Rollup Stack Apart from a Single L2
| Quick answer: A rollup stack is an open-source, modular software framework that provides the shared infrastructure for deploying blockchain networks, whereas a Layer 2 is one specific network that runs on top of that framework. |
Think of it this way: the stack is the operating system. The L2 is an application that runs on it.
Three frameworks currently dominate the L2 deployment landscape:
Framework | Developer | Proof System | Key Trait |
| OP Stack | OP Labs / Optimism | Optimistic (fault proofs) | Ethereum-aligned, Superchain interop |
| Arbitrum Orbit | Offchain Labs | Optimistic (Nitro + BoLD) | Maximum customization, L3 strategy |
| ZK Stack | Matter Labs / zkSync | ZK (validity proofs) | Cryptographic finality, no fraud window |
These three differ across proof systems, EVM compatibility surface, withdrawal latency, data availability options, and ecosystem model, but all share the same underlying purpose: making it easier for teams to spin up their own rollup without starting from scratch.
Rollup Stacks vs L2s: Similar Goals, Different Roles
| At a glance: Both rollup stacks and individual L2s aim to scale Ethereum, but a rollup stack sets the architectural rules (proof system, data availability, interoperability), while an L2 applies those rules to serve a specific user base or use case. |
- An individual L2 makes decisions about user experience: which tokens to accept as gas, how fast blocks are produced, what governance model to use, and which applications to attract.
- A rollup stack makes architecture decisions: which proof system to use, how data is posted to Ethereum, how chains within the ecosystem communicate with each other, and how security upgrades are coordinated.
In practical terms:
- The stack determines whether two chains can talk to each other natively.
- The L2 determines what those chains are used for.
Base and OP Mainnet are both distinct networks with different users, liquidity pools, and sequencer policies. But they run on the same OP Stack, which means they share the same proof logic, the same upgrade governance, and the same path toward Superchain interoperability.
Arbitrum One and a hypothetical Robinhood L2 (which Robinhood has announced it is building on Arbitrum Orbit) are similarly distinct chains that inherit the same Nitro security stack and tooling.
The stack is the shared foundation. The L2 is just one expression of it.
Can Different Layer 2s Use the Same Rollup Stack?
| Quick answer: Yes. Multiple independent L2s can run on the same rollup stack, sharing its security model, developer tooling, and interoperability infrastructure, while remaining entirely separate networks with different users, governance, and use cases. This is precisely what makes rollup stacks powerful. |
The OP Stack is the clearest example.
As of 2025, the OP Stack powers more than 40 production chains, including Base (Coinbase's L2), OP Mainnet, Zora, Mode, Redstone, Unichain (Uniswap's L2), and Kraken's Ink. Together, these chains process several hundred million dollars in daily on-chain volume.
Arbitrum Orbit follows a similar pattern. Arbitrum chains are configurable instances of the Arbitrum Nitro tech stack. Teams can adjust execution, fee models, governance, data availability, and validation parameters for their specific use case.
This model has a significant implication: when a stack succeeds, all chains on it benefit. Security upgrades, new DA layer integrations, and interoperability features roll out across every chain using that framework, not just the flagship L2.
It also works in reverse. If a stack has a critical vulnerability, every chain on it is exposed. Choosing a stack is a risky decision.
Why Rollup Stacks May Matter More Than Individual L2s
| Quick answer: Because fees across all L2s have converged to near-zero after EIP-4844, the real competitive differentiator is no longer cost. It is ecosystem composability, interoperability, and shared security upgrades. All three are determined at the stack level, not the chain level. |
After EIP-4844 (Dencun upgrade, March 2024) cut L2 data-posting costs by 80–90%, per-transaction fees fell below $0.10 across every major network. The fee gap between chains collapsed.
When fees converge, the competition shifts to something harder to replicate: ecosystem lock-in, cross-chain composability, and shared upgrade governance – all of which are determined at the stack level.
Consider the data as of early 2026:
- Arbitrum One holds approximately 44% of the total L2 TVL.
- Base (OP Stack) holds approximately 33%.
- Together, chains running on OP Stack and Arbitrum Orbit account for the overwhelming majority of all L2 DeFi liquidity.
Base did not grow because it had cheaper fees than Arbitrum. It grew because Coinbase's distribution infrastructure, combined with OP Stack's shared security and Superchain interoperability, created a compounding advantage that no single-chain strategy could match.
My own perspective
What becomes clear when you look past chain-level metrics is that the real competition in Ethereum scaling has already moved to the stack layer, and most market participants haven't noticed yet. Tracking which L2 has the highest TVL this week is equivalent to tracking which app has the most downloads without asking what operating system it runs on. The stack determines the ceiling: how interoperable a chain can be, how quickly it inherits security improvements, and whether its liquidity is trapped or composable. Builders who chose OP Stack in 2023 plugged into a network that would eventually include Base's 100M+ potential users. That is a stack-level bet, not a chain-level one.
Rollup Stacks Are Becoming Ethereum's Real Battleground
| At a glance: The competition is no longer Base vs. Arbitrum. It is OP Stack ecosystem vs. Arbitrum Orbit ecosystem vs. ZK Stack ecosystem, each pulling builders and liquidity into a different long-term orbit within Ethereum's rollup ecosystem. |
- OP Stack prioritizes Ethereum alignment and standardization. It operates under an MIT license (with a 2.5% revenue share or 15% of on-chain profit for chains joining the Superchain). The Superchain registry lists over 40 chains that have opted into shared upgrade governance.
- Arbitrum Orbit prioritizes flexibility and sovereignty. It supports both optimistic rollup and AnyTrust mode (using a Data Availability Committee for reduced costs), and its Stylus VM enables smart contracts written in Rust, C, and C++ alongside Solidity. Orbit chains can deploy as L2s (settling to Ethereum) or L3s (settling to Arbitrum One). The tradeoff: independent Orbit chains require a 10% profit share with the Arbitrum Foundation.
- ZK Stack offers cryptographic rollup finality without a fraud window. Validity proofs verify every batch before it is accepted on Ethereum. This eliminates the 7-day withdrawal delay inherent to optimistic rollups, but introduces higher prover infrastructure costs.
Each stack involves a trade between sovereignty and ecosystem benefits, and each locks builders into a self-serving ecosystem. The critical insight is that this lock-in is increasingly intentional. Builders are choosing stacks for the network of chains they will be able to interoperate with.
What This Means for Developers and Builders
| Quick answer: For developers, choosing a rollup stack is a founding-level decision that determines interoperability, licensing cost, available tooling, and switching cost, not a configuration choice that can be easily reversed later. |
The stack you choose determines:
- Interoperability: Which chains can you natively compose with, without third-party bridges?
- Security inheritance: How do security upgrades propagate to your chain?
- Licensing and cost: What revenue do you owe the stack's governance body?
- Developer tooling: Which languages, VMs, and infrastructure are available out of the box?
Choose OP Stack if you want Ethereum alignment, Superchain interoperability, an MIT-licensed open-source codebase, and access to an ecosystem of 40+ chains.
Choose Arbitrum Orbit if you need maximum customization — custom gas tokens, AnyTrust DA, Stylus for non-EVM languages, or an L3 architecture that settles to Arbitrum One rather than Ethereum directly.
Choose ZK Stack if cryptographic finality is non-negotiable for your use case (institutional finance, high-value settlements) and you can absorb prover infrastructure costs.
Switching stacks after launch is not a configuration change. It means rebuilding integrations, retraining teams, redeploying contracts, and asking users and liquidity to migrate with you. Stack choice is closer to a founding decision than a technical preference.
What This Means for Investors and Ecosystem Watchers
| Quick answer: For investors, the more meaningful signal is not which L2 has the highest TVL today. It is the rollup stack that is attracting the most new chains, because that determines where builders, liquidity, and users will compound over time. |
When a new chain launches on OP Stack, it expands the surface area of the entire Superchain ecosystem – more potential liquidity, more composable protocols, more users that Base or OP Mainnet can interact with natively.
This has direct implications for token holders:
- OP token derives value not just from OP Mainnet's activity, but from the growth of every chain in the Superchain ecosystem that contributes to Optimism's collective governance and revenue model.
- ARB token similarly reflects the health of the entire Orbit ecosystem, including the L3 chains that settle on Arbitrum One.
The metrics worth tracking are the number of new chains deploying on each stack, the total combined TVL across all chains on the stack, and cross-chain transaction volume as an indicator of actual composability in use.
As of 2026, 73 active rollups collectively secure more than $48 billion in TVL, roughly double the figure from early 2025. But that growth is concentrating around a small number of stacks. The stack winners are already visible. What remains uncertain is how wide the gap becomes.
Sources and Further Reading
- Ethereum.org – "Layer 2 Scaling" https://ethereum.org/en/layer-2/
- Vitalik Buterin – "A Rollup-Centric Ethereum Roadmap" https://ethereum-magicians.org/t/a-rollup-centric-ethereum-roadmap/4698
- Optimism Docs – "OP Stack Introduction" https://docs.optimism.io/stack/getting-started
- Arbitrum Docs – "Arbitrum Chains: A Gentle Introduction" https://docs.arbitrum.io/launch-arbitrum-chain/a-gentle-introduction
- QuickNode – "Which Rollup Framework Should You Use?" https://www.quicknode.com/guides/custom-chains/which-rollup-framework-should-you-use-for-your-rollup
- Gelato – "OP Stack vs Arbitrum Orbit: The Best L2 Rollup Comparison" https://gelato.cloud/blog/op-stack-vs-arbitrum-orbit-the-best-l2-rollup-comparison
FAQs About Rollup Stacks Vs L2s
Technically, yes. Early L2s like the original Optimism were built as custom, monolithic codebases. But in practice, virtually all new L2s launched since 2023 use an established rollup framework. Building from scratch offers full sovereignty but requires maintaining a divergent codebase indefinitely, with no shared security upgrade path.
