• Sequencer Commitment Games (SCGs) give OP Stack rollups instant, trust-minimized transaction execution guarantees.

• No protocol changes needed: Integration via a lightweight sequencer side-car, without modifications to the rollup or Ethereum L1 core logic, enabling immediate adoption by existing OP Stack chains without forks or governance delays.

• Major UX upgrades: Delivers instant trade confirmations, trust-minimized bridging, and sub-10 ms inclusion guarantees. Reduces multi-minute or multi-day waits to near-Web2 speed for transactions and bridging.

• Bundle support: SCGs add cryptoeconomically protected support of transaction bundles to all participating rollups, enabling advanced mev strategies.

• Interoperability boost: Provides seamless coordination between OP Stack rollups and Ethereum L1 through mev-commit’s live PBS actors on mainnet. Supports cross-domain mev strategies and stateful workflows.

• Cross-rollup composability: Enables atomic execution across multiple OP Stack chains and L1, and other rollups that integrate mev-commit, unlocking the potential for cross-rollup dApps and efficient liquidity routing.

• Economic upside: Creates new revenue streams from preconfirmation fees and cross-domain coordination incentives, benefiting sequencers and validators. This will likely also lead to an increase in TVL.

• Ecosystem leadership: Potential to position the OP Stack as the first rollup ecosystem to interoperate with Ethereum mainnet and standardize trust-minimized chain coordination, setting a precedent for the broader Ethereum landscape.

The rise of Layer 2 scaling solutions marks a pivotal moment in Ethereum’s evolution. Among these, OP Stack-based chains such as Optimism, Base, and others have gained prominence for their modularity, open-source design, and vision of a “Superchain” composed of interoperable rollups. These chains promise to alleviate Ethereum’s congestion and high gas fees by offloading execution to secondary layers, while maintaining security through L1 settlement. However, as the Layer 2 ecosystem matures, it has become increasingly clear that significant systemic challenges remain unresolved.

One of the most pressing issues is liquidity fragmentation. As capital spreads across a growing number of rollups, each with its own bridge infrastructure and token ecosystem, liquidity becomes siloed. This not only reduces capital efficiency but also creates a poor user experience: bridging assets between rollups or from L1 to L2 often requires multiple transactions, long wait times, or exposure to trust assumptions in third-party bridges. Overall, there is a lack of native interoperability, both between Layer 2 rollups and between rollups and Ethereum L1. The result is a fragmented economic landscape where dApps struggle to achieve composability and users are penalized for navigating across chains.

In addition to liquidity silos, many OP Stack chains still suffer from slow or inconsistent confirmation times, particularly in cross-domain interactions. Although optimistic rollups offer cheaper and scalable execution, many OP Stack chains have a block time of 2 seconds. While this is faster than L1 block times, it is still a far cry from a Web2 UX. Even worse, for transaction finality, often required for bridging, users have to wait a multi-day challenge period.

To deliver a single‑chain user experience, which is the stated goal of the Optimism Superchain initiative, these shortcomings must be resolved. We propose Sequencer Commitment Games (SCGs) as a solution. Building on the production‑grade mev‑commit network, an SCG lets a sequencer escrow collateral and sign a binding statement about how and when it will handle a transaction bundle. If the statement proves false, it is reported by an oracle and the collateral gets slashed. The mechanism operates on the mev-commit chain separate from the Optimism protocol and therefore requires no hard‑fork or protocol modifications. Among other use-cases this unlocks instant bridging both between rollups and L1 as well as between different rollups with sub-second instead of multi-day latency. Furthermore, mev-commit allows searchers to submit transaction bundles to rollups with the cryptoeconomic guarantee that these bundles are not broken up. The combined benefits will likely increase the usage of participating OP stack chains and consequently lead to an increase in TVL. This article describes SCGs in detail and explains how they can benefit the OP stack.

Ethereum’s Proposer–Builder Separation (PBS) divides the responsibility of block building into two distinct roles. Block builders assemble candidate payloads and attach a bid that represents the amount of ETH they are willing to pay the slot’s proposer for proposing their block. The proposer, who sees only a hash of each payload, selects the highest bid and forwards the winning header to the consensus layer. From the perspective of builder ↔ proposer coordination, PBS succeeds by maximizing validator revenue and reducing the need for proposers to run sophisticated MEV infrastructure. Yet PBS lacks user‑level guarantees: A block builder is free to reorder and censor transactions.

The concept of commitment games has been introduced by Primev in the article “Commitment Games and Where to Find Them” [1]. A commitment game is a cryptographic engagement between a bidder and an execution service provider, backed by stake and enforced by an oracle. Although the concept is general, the setting that concerns us here involves exactly two parties:

• The bidder (e.g., a user, wallet, searcher or liquidity router) proposes a condition (or fulfillment predicate) on a future state: for example, “include transaction $$T$$ in block height $$n + 1$$ on chain $$C$$,” or “ensure that bundle $$B$$ executes on chain $$C$$ without reverting.” The bid also names (i) a reward $$f$$ to be paid if the condition is met, and (ii) a penalty $$p$$ to be paid to the bidder if the provider deviates.

Once the outcome of the condition is determined, e.g., the block at height $$n+1$$ has been finalized, the commitment is opened and a deterministic oracle checks the condition. If it is satisfied, the reward $$f$$ is paid to the provider; otherwise, the penalty $$p$$ is taken from the provider’s collateral to compensate the bidder. As long as the values $$f$$ and $$p$$ are chosen appropriately, deviation costs the provider more than it can gain and therefore the commitment is economically credible.

A live instantiation of commitment games on Ethereum L1 mainnet and the Hoodi testnet is available today by the mev-commit network, which is described in more detail in the following section.

Preconfirmations or preconfs for short as introduced by Primev in a Mirror article [2] are one of the simplest special cases of commitment games. The fulfilment predicate is inclusion (and optional successful execution) of a single transaction or a transaction bundle in a designated future block. The providers for preconfs can be the block proposers or the block builders for Ethereum L1 preconfs, or the sequencer for L2 preconfs. Preconfs for L1 with block builders as commitment providers are live via mev-commit today.

A more advanced type of commitment are synchronous commitments, in which several providers for different chains participate. A synchronous commitment is only given if all participants agree to commit. The arrangement enables atomic execution across chains without any one provider trusting the others, because failure to comply triggers slashing for every deviator. Synchronous commitments extend the design space for the participating rollup ecosystems and unlock massive value by enhancing UX, increasing liquidity, and by enabling mainnet interoperability and efficient cross-chain operations.

The mev-commit network has been developed by Primev following early discussions with the Ethereum Foundation and Flashbots at Devcon Bogotá 2022. It was designed to support fast, private, and credibly neutral coordination across diverse actors in the MEV supply chain by enabling commitment games. The protocol operates through a dedicated coordination layer, the mev-commit chain, which is used to record and settle commitments in near real-time. All components of the mev-commit network are open-source (all code is available on GitHub). The protocol is chain-agnostic and supports any application that benefits from strong inclusion or execution guarantees, including bridges, searchers, rollups, and intent-based systems.

The mev-commit protocol enables commitment games as defined above. In its current deployment, it focuses on preconfirmations with Ethereum block builders as commitment providers. Bidders can specify in their bids which transactions, or transaction bundles, they want to have included at a certain block height. Furthermore, bidders can specify whether they only want inclusion guarantees, or whether they additionally want a commitment for the transactions executing without reverting. These commitments are posted to the mev-commit chain, an EVM chain with ~200ms block times (and as low as 5ms on testnet), optimized for high-throughput commitment settlement. Whenever a block builder builds the corresponding L1 block, this builder’s commitments will be checked for fulfillment and rewards or penalties will be paid between the bidders and this block builder. Commitments offer cryptographically secure privacy and their timeliness is incentivized via a decay mechanism.

We provide a high-level overview of the protocol below. For more detailed information, we refer to the mev-commit whitepaper [3] and the mev-commit documentation [4]. See Figure 1 below (taken from the whitepaper [3]) for a depiction of the basic interaction between the involved actors.

Privacy is essential to prevent strategic manipulation, frontrunning, or value leakage in MEV markets. To this end, mev-commit is designed with end-to-end privacy as a first-class feature. The protocol enforces three privacy guarantees:

1. Bid Privacy: Only intended providers can read the contents of a bid.

2. Commitment Privacy: Commitment contents remain hidden until explicitly settled.

3. Recipient Anonymity: Observers cannot determine which providers received a bid until it is revealed.

To achieve this, mev-commit introduces two cryptographic primitives:

• Anonymous Receiver-Updatable Broadcast Encryption (ARUBE) enables bidders to send encrypted bids to a dynamic set of providers without revealing the recipient list. This prevents attackers from deducing intent or frontrunning based on bid flow. It also provides recipient anonymity.

• Dual-Phase Commitments (DPCOM) allow both the bidder and provider to open a commitment, ensuring mutual accountability while maintaining confidentiality until settlement. If a provider fails to reveal a fulfilled commitment, the bidder can do so unilaterally.

These primitives combine to deliver a protocol that supports private coordination at scale, an essential requirement for use cases like private searcher flows, cross-chain bridging, and latency-sensitive applications. More information about privacy can be found in an article dedicated to it [5].

A key challenge in preconfirmation protocols is ensuring that providers issue commitments before the corresponding transaction is confirmed on-chain. Without safeguards, a provider could opportunistically wait to see a transaction included and then issue a "preconfirmation" retroactively, undermining its credibility. This issue is known as the preconfirmation fair exchange problem [6, 7].

mev-commit addresses this with a time-decay bidding mechanism. Each bid specifies a value function $$\beta(t)$$ that decreases linearly over time, starting from the moment the bid is issued and decaying to zero by a fixed expiry time. This creates strong cryptoeconomic pressure for providers to commit as early as possible, in particular before the relevant block is built. Late commitments receive reduced or no payment, making opportunistic behavior economically unviable.

The decay is enforced via timestamps recorded on the mev-commit chain, and validated using a timing tolerance parameter, which accounts for message latency. This ensures that honest providers with reasonable network conditions can participate without penalty, while discouraging timestamp manipulation.

By embedding decay directly into bid valuation, mev-commit solves the fair exchange problem in a decentralized setting, enabling credible commitments without trusted third parties.

The opt-in mechanism in mev-commit allows Ethereum validators to assert to only proposing blocks built by registered mev-commit builders. Validators signal their participation by registering their BLS public key either via a direct opt-in or via restaking through Eigenlayer or Symbiotic. In a rollup setting, this role can be played by sequencers as well as validators.

Opt-in plays a critical role in strengthening the reliability and enforceability of commitments: Users and bidders can place greater trust in commitments whenever the current validator is opted in, as the risk of non-inclusion due to proposing a block from a builder outside of mev-commit is mitigated. Furthermore, validators benefit economically from opting in by enabling higher-value blocks from builders who have received mev-commit bids. This increases builder competitiveness in mev-boost auctions and indirectly boosts proposer rewards.

mev-commit has seen substantial adoption across Ethereum’s consensus and execution layers. As of writing, more than 10,000 validators have opted into the protocol, signaling willingness to support blocks constructed by mev-commit registered builders.

Integration is also widespread across the builder and relay ecosystem. A majority of active block builders and relays support mev-commit, ensuring that a large portion of Ethereum’s blockspace can be coordinated through the protocol. This high level of builder participation enables bidders to source commitments from multiple providers, significantly increasing the probability that their desired execution paths are fulfilled.

This combination of builder coverage and validator opt-in ensures that mev-commit can offer near-instant execution assurances with strong economic backing.

Empirical data published on the Ethereum Research forum [8] document 807 execution preconfirmations across 415 Holesky blocks. Importantly, block value doubled: mev‑commit blocks averaged 0.0093 ETH versus 0.0044 ETH for vanilla counterparts. The figures confirm that commitment bids can significantly contribute to the overall block value. Mainnet growth continues developing, and can be tracked on the mev-commit Xplorer and this Dune Dashboard.

An Optimism sequencer wishing to issue commitments first stakes collateral on the mev-commit chain and registers its commitment key. It then launches a lightweight provider side‑car alongside the existing node. The side‑car’s duties are purely at the networking layer: receive bids labelled with the roll‑up’s chain‑ID, decrypt them, evaluate their fees, and, upon accepting, sign and send commitments. Importantly, no modification to the consensus or execution binaries is necessary. The sequencer only needs to ensure that commitments are only issued if fulfillment can be guaranteed, e.g., after already internally queuing the corresponding transactions for inclusion in the next block.

Note that mev-commit already works with multiple block builders on Ethereum L1. Using the same techniques as there, it is therefore also compatible with decentralized sequencers, where each node can run a provider node. The approach is thus future proof.

To enable synchronous commitments and atomic transaction bundles across different chains, a deeper integration is necessary. We here describe different alternatives how this could be achieved and their tradeoffs. The first approach via a dedicated synchronous commitment protocol provides the largest flexibility, but also requires all participants to engage in the protocol. As lightweight alternatives, we describe two related methods that still achieve some of the desired functionality with little extra effort from the commitment providers.

Synchronous Commitment Protocol. A bundle consisting of transactions for different chains requires coordination between the commitment providers of all these chains. For rollups, the providers are the respective sequencers and for Ethereum L1 the providers are all mev-commit registered block builders. A synchronous commitment should be issued if and only if all these providers agree on fulfilling their part of the transaction bundle (and if the relevant validator is opted in for L1 transactions), because only then atomic execution across all chains can be guaranteed.

Agreement either to include all transactions or none can be reached among the providers via a consensus protocol, either executed directly among themselves, or leveraging the mev-commit chain. Once consensus is reached, the corresponding commitment can be stored on the mev-commit chain. As for regular commitments, in case the synchronous commitment gets violated, the deviating providers get slashed and the corresponding bidder gets compensated.

Atomic Cross‑Chain Execution via State‑Lock Commitments. If some of the involved providers are not willing or able to participate in the consensus protocol described above, but instead can issue state-lock commitments, most of the functionality can still be achieved. This method has been suggested by AJ Park in the Flashbots research forum [9]. The basic idea is as follows: Bidders can obtain a state-lock commitment, granting the bidder exclusive access to a specified state (e.g., a specific Uniswap pool) for the next block. That is, after such a commitment is issued, transactions touching that state from everybody except the commitment’s bidder must be rejected, and transactions from the bidder must be included.

Given such state-lock commitments, bidders can coordinate atomic cross-chain bundles themselves as follows: First, bid for state-lock commitments for the relevant states on all involved chains. If such commitments are obtained from all chains, submit the corresponding transactions to all the chains. In that case, execution on all chains is guaranteed. If no state-lock commitment can be obtained on at least one chain in a timely manner, do not submit any of the involved transactions and potentially retry for the next block (or consider the bundle failed).

This method ensures that either all transactions in the bundle get executed on all chains, or none of them get executed (or the bidder gets compensated by a provider violating a commitment).

Atomic Cross‑Chain Execution via Cancellable Commitments. If state-lock commitments are not supported by all providers, a similar approach can be used via cancellable commitments: These are commitments that can be cancelled up to some deadline and if cancelled, the corresponding transactions must not be included. This method can even be employed if not all providers support cancellations. Consider, e.g., the scenario where cancellations are supported on L1, but not on some rollup and a user wants atomic execution across this rollup and L1. In that case, the user can first try to obtain a (cancellable) commitment from all providers on L1. If all these commitments are obtained, try to get a commitment on the rollup. If this is also obtained, all transactions are guaranteed to get executed (or the the bidder gets compensated). If no commitment can be obtained on the rollup, cancel the L1 commitments. In that case, no transaction will get executed, again ensuring atomic execution.

Bridging between Ethereum L1 and OP Stack rollups, or between different OP Stack chains, typically suffers from latency up to multiple days. The bottleneck is the time it takes for the origin chain transaction depositing funds into the bridge to reach confirmation or finality before the destination can release the funds. Sequencer Commitment Games (SCGs) remove this bottleneck by letting the origin chain’s block producer (L1 builder or L2 sequencer) issue an economically enforceable preconfirmation via the mev-commit network.

This source-chain preconfirmation cryptoeconomically guarantees that the bridge-initiating transaction will be included in a future block. Once a bridge operator or liquidity provider holds this guarantee, they can release funds instantly on the destination chain, because either (a) the origin transaction will appear as committed, or (b) the provider will be compensated from the violator’s collateral if it does not.

With this approach, instant bridging is possible in all directions between Ethereum L1 and different L2s. The pattern is the same for all directions:

• L1 → L2: L1 block builder commits to include the deposit transaction; destination L2 credits instantly.

• L2 → L1: L2 sequencer commits to include the withdrawal initiation; destination L1 credits instantly.

• Cross-rollup: Source rollup sequencer commits; destination rollup credits instantly.

See Figure 2 below for an overview of the general flow for an instant bridging transaction.

In today’s deployment, mev-commit preconfirmations are available for Ethereum L1 transactions, so instant bridging is live for L1-origin flows. Extending the same guarantees to L2-origin transfers requires integrating L2 sequencers as registered commitment providers.

The integration of Sequencer Commitment Games (SCGs) into OP Stack rollups provides a suite of technical, user-facing, and ecosystem-level advantages that address persistent challenges in scalability, interoperability, and liquidity.

By enabling low-latency preconfirmations, SCGs allow sequencers to issue economically credible inclusion or execution guarantees within as little as 5 ms. For traders, DeFi protocols, and wallets, SCGs enable instant trade confirmations, with execution guaranteed or compensated if deviations occur. This reduces transaction settlement uncertainty to near real-time and brings Layer 2 UX close to Web2 responsiveness.

As an additional benefit, mev-commit enables searchers to submit transaction bundles in the form of bids to rollups, a feature currently not supported natively by most rollups. In addition to simply submitting transaction bundles, which mev-searchers commonly do when interacting with L1 block builders, bundles via SCGs are economically protected: once committed to, their inclusion is backed by collateral, and any deviation triggers slashing. This means searchers get the guarantee that all transactions in the bundle get included in the same order as they appeared in the bundle and no other transactions will get in between, enabling complex searcher strategies.

Through integration with the mev-commit network and its pool of opted-in PBS actors on Ethereum mainnet, OP Stack rollups gain native interoperability with other rollups and Ethereum L1 without introducing new trust assumptions. Commitments can be coordinated with other sequencers and L1 block builders to enable seamless cross-domain workflows. Synchronous commitments can be used to enable atomic multi-chain execution. This increases capital efficiency by enabling sophisticated strategies such as atomic cross-chain arbitrage. As a related benefit, SCGs facilitate instant bridging between rollups and between L1 and L2, bypassing multi-minute to multi-day bridging delays. The resulting UX is trust-minimized, removing reliance on external bridge operators. All guarantees are enforced by the same cryptoeconomic security model as the underlying commitments.

The advantages are summarized in the table below, comparing the status quo with what can be achieved via SCGs:

Aspect                       | Status Quo             | With SCGs
-----------------------------|------------------------|-----------------
Transaction confirmation     | Up to 2 seconds        | As fast as 5 ms
L1 <-> Rollup Bridge Latency | Minutes to days        | Sub-seconds
MEV Bundle Support           | No                     | Yes
Validator/Relay Engagement   | No                     | Yes
Sequencer Modifications      | Required for upgrades  | None
UX Parity with Web2          | No                     | Much closer

A neutral, chain-abstracted commitment layer such as mev-commit creates new structural advantages for the OP Stack ecosystem, which are likely to increase TVL:

• Better liquidity routing: Commitments make liquidity across L1 and L2s more fungible by reducing settlement times, drastically reducing bridging delays, and enabling atomic execution, allowing to allocate capital where it earns the highest yield in real time.

• Support for cross-chain dApps: Developers can build applications that rely on fast, guaranteed execution across multiple rollups and L1, unlocking use cases that were previously infeasible due to settlement risk or latency.

• Shared revenue layer: Relays, sequencers, and validators participating in the commitment layer can share in new revenue streams generated by preconfirmation fees and cross-domain transaction flows.

The OP Stack is well-positioned to adopt SCGs early: its modular architecture, shared vision of the “Superchain,” and centralized-sequencer deployment model provide a straightforward integration path without protocol-level changes. Furthermore, new OP Stack rollups can onboard smoothly: once a sequencer integrates SCGs, it can immediately interoperate with Ethereum mainnet and other OP Stack chains through the shared commitment layer, without building bespoke infrastructure. To integrate, no customization of the sequencer logic is required and mev-commit is compatible with all kinds of sequencers, including those running inside a TEE.

Sequencer Commitment Games (SCGs) offer a practical, high-impact path to improving the UX, interoperability, and capital efficiency of the OP Stack ecosystem, without requiring any core protocol changes. By building on the proven mev-commit network, SCGs bring economically enforceable inclusion and execution guarantees to OP Stack sequencers, enabling low-latency preconfirmations, synchronous cross-chain coordination, and instant bridging between L1 and L2 as well as across rollups.

From a user perspective, SCGs deliver near-instant trade confirmations, trust-minimized bridging, and a Web2-like transaction experience. For searchers and liquidity providers, they unlock new mev strategies on L2s with cryptoeconomic protection against bundle breakage. They integrate seamlessly with Ethereum mainnet’s opted-in PBS actors, enabling native interoperability between rollups and L1.

At the ecosystem level, a neutral, chain-abstracted commitment layer can unify liquidity routing, support cross-chain dApps, and create shared revenue streams for participating actors, likely increasing TVL. The OP Stack is uniquely positioned to lead this effort: its modular design, Superchain vision, and sequencer architecture make adoption straightforward, allowing new rollups to plug into a shared commitment infrastructure from day one.

Key highlights include:

• No protocol-level changes: Integration happens via a side-car; sequencer logic stays untouched.

• Major UX gains: Instant confirmations and bridging, cross-rollup composability.

• Interoperability upside: Trust-minimized coordination with Ethereum L1 and other L2s.

• OP ecosystem leadership: Opportunity to set the standard for cross-chain coordination.

The research and development of SCGs for OP Stack rollups is at a pivotal stage. To realize this vision, collaboration across sequencer operators, rollup teams, wallets, and dApp developers is essential. We encourage the OP Stack community, core contributors, governance participants, and ecosystem partners to engage with the design process, test integrations, and provide feedback. With early adoption, the OP ecosystem can pioneer a new era of trust-minimized, chain-coordinated infrastructure and set a precedent for the broader Ethereum rollup landscape.

If supported by Optimism, Primev intents to continue working on the integration, in particular:

• Integration with Superchain sequencers, enabling synchronous commitments across the entire OP Stack ecosystem.

• Native wallet and provider tooling support, allowing users to request and verify commitments directly from their transaction interface.

• DAO onboarding of PBS actors, aligning governance with the economic incentives of block builders, relays, and sequencers to maximize adoption.

Acknowledgements. This research was supported by a grant from Optimism Foundation.

[1] Primev. Commitment Games and Where to Find Them, Mirror, 11 Jun 2024. https://mirror.xyz/preconf.eth/3SXageNPpcQMz8L2YfW5ipvthNl-InPzaPAfb3LxwqA

[2] Primev. Preconfirmations: The Fulfillment-Delivery Paradigm, Mirror, 7 Feb 2024. https://mirror.xyz/preconf.eth/sgcuSbd1jgaRXj9odSJW-_OlWIg6jcDREw1hUJnXtgI

[3] Primev. mev‑commit Whitepaper, v 1.0, 25 Nov 2024. https://whitepaper.mev-commit.xyz/

[4] Primev. mev-commit documentation. https://docs.primev.xyz/

[5] Primev. Pioneering End-to-End Privacy for Intra-Block Commitments, Mirror, 18 Sep 2024. https://mirror.xyz/preconf.eth/iz2J0uTXHhl8DiAkG-VLLwvCp-8qcc_Z7A8_4rU0A3g

[6] Justin Drake. Based preconfirmations. Ethereum Research, 2023. https://ethresear.ch/t/based-preconfirmations/.

[7] Yaonam. Solutions to the preconf fair exchange problem. Ethereum Research, 2024. https://ethresear.ch/t/solutions-to-the-preconf-fair-exchange-problem/.

[8] Evan Kim. Preconfirmation Bidding Increased Block Values on Holesky, ethresear.ch, 2024. https://ethresear.ch/t/preconfirmation-bidding-increased-block-values-on-holesky/

[9] AJ Park. Atomic Cross‑Rollup Arbitrage via Coordinator and State‑Lock Auctions, The Flashbots Collective, 12 Jun 2025. https://collective.flashbots.net/t/atomic-cross-rollup-arbitrage-via-coordinator-and-state-lock-auctions/

• Sequencer Commitment Games (SCGs) give OP Stack rollups instant, trust-minimized transaction execution guarantees.

• No protocol changes needed: Integration via a lightweight sequencer side-car, without modifications to the rollup or Ethereum L1 core logic, enabling immediate adoption by existing OP Stack chains without forks or governance delays.

• Major UX upgrades: Delivers instant trade confirmations, trust-minimized bridging, and sub-10 ms inclusion guarantees. Reduces multi-minute or multi-day waits to near-Web2 speed for transactions and bridging.

• Bundle support: SCGs add cryptoeconomically protected support of transaction bundles to all participating rollups, enabling advanced mev strategies.

• Interoperability boost: Provides seamless coordination between OP Stack rollups and Ethereum L1 through mev-commit’s live PBS actors on mainnet. Supports cross-domain mev strategies and stateful workflows.

• Cross-rollup composability: Enables atomic execution across multiple OP Stack chains and L1, and other rollups that integrate mev-commit, unlocking the potential for cross-rollup dApps and efficient liquidity routing.

• Economic upside: Creates new revenue streams from preconfirmation fees and cross-domain coordination incentives, benefiting sequencers and validators. This will likely also lead to an increase in TVL.

• Ecosystem leadership: Potential to position the OP Stack as the first rollup ecosystem to interoperate with Ethereum mainnet and standardize trust-minimized chain coordination, setting a precedent for the broader Ethereum landscape.

The rise of Layer 2 scaling solutions marks a pivotal moment in Ethereum’s evolution. Among these, OP Stack-based chains such as Optimism, Base, and others have gained prominence for their modularity, open-source design, and vision of a “Superchain” composed of interoperable rollups. These chains promise to alleviate Ethereum’s congestion and high gas fees by offloading execution to secondary layers, while maintaining security through L1 settlement. However, as the Layer 2 ecosystem matures, it has become increasingly clear that significant systemic challenges remain unresolved.

One of the most pressing issues is liquidity fragmentation. As capital spreads across a growing number of rollups, each with its own bridge infrastructure and token ecosystem, liquidity becomes siloed. This not only reduces capital efficiency but also creates a poor user experience: bridging assets between rollups or from L1 to L2 often requires multiple transactions, long wait times, or exposure to trust assumptions in third-party bridges. Overall, there is a lack of native interoperability, both between Layer 2 rollups and between rollups and Ethereum L1. The result is a fragmented economic landscape where dApps struggle to achieve composability and users are penalized for navigating across chains.

In addition to liquidity silos, many OP Stack chains still suffer from slow or inconsistent confirmation times, particularly in cross-domain interactions. Although optimistic rollups offer cheaper and scalable execution, many OP Stack chains have a block time of 2 seconds. While this is faster than L1 block times, it is still a far cry from a Web2 UX. Even worse, for transaction finality, often required for bridging, users have to wait a multi-day challenge period.

To deliver a single‑chain user experience, which is the stated goal of the Optimism Superchain initiative, these shortcomings must be resolved. We propose Sequencer Commitment Games (SCGs) as a solution. Building on the production‑grade mev‑commit network, an SCG lets a sequencer escrow collateral and sign a binding statement about how and when it will handle a transaction bundle. If the statement proves false, it is reported by an oracle and the collateral gets slashed. The mechanism operates on the mev-commit chain separate from the Optimism protocol and therefore requires no hard‑fork or protocol modifications. Among other use-cases this unlocks instant bridging both between rollups and L1 as well as between different rollups with sub-second instead of multi-day latency. Furthermore, mev-commit allows searchers to submit transaction bundles to rollups with the cryptoeconomic guarantee that these bundles are not broken up. The combined benefits will likely increase the usage of participating OP stack chains and consequently lead to an increase in TVL. This article describes SCGs in detail and explains how they can benefit the OP stack.

Ethereum’s Proposer–Builder Separation (PBS) divides the responsibility of block building into two distinct roles. Block builders assemble candidate payloads and attach a bid that represents the amount of ETH they are willing to pay the slot’s proposer for proposing their block. The proposer, who sees only a hash of each payload, selects the highest bid and forwards the winning header to the consensus layer. From the perspective of builder ↔ proposer coordination, PBS succeeds by maximizing validator revenue and reducing the need for proposers to run sophisticated MEV infrastructure. Yet PBS lacks user‑level guarantees: A block builder is free to reorder and censor transactions.

The concept of commitment games has been introduced by Primev in the article “Commitment Games and Where to Find Them” [1]. A commitment game is a cryptographic engagement between a bidder and an execution service provider, backed by stake and enforced by an oracle. Although the concept is general, the setting that concerns us here involves exactly two parties:

• The bidder (e.g., a user, wallet, searcher or liquidity router) proposes a condition (or fulfillment predicate) on a future state: for example, “include transaction $$T$$ in block height $$n + 1$$ on chain $$C$$,” or “ensure that bundle $$B$$ executes on chain $$C$$ without reverting.” The bid also names (i) a reward $$f$$ to be paid if the condition is met, and (ii) a penalty $$p$$ to be paid to the bidder if the provider deviates.

Once the outcome of the condition is determined, e.g., the block at height $$n+1$$ has been finalized, the commitment is opened and a deterministic oracle checks the condition. If it is satisfied, the reward $$f$$ is paid to the provider; otherwise, the penalty $$p$$ is taken from the provider’s collateral to compensate the bidder. As long as the values $$f$$ and $$p$$ are chosen appropriately, deviation costs the provider more than it can gain and therefore the commitment is economically credible.

A live instantiation of commitment games on Ethereum L1 mainnet and the Hoodi testnet is available today by the mev-commit network, which is described in more detail in the following section.

Preconfirmations or preconfs for short as introduced by Primev in a Mirror article [2] are one of the simplest special cases of commitment games. The fulfilment predicate is inclusion (and optional successful execution) of a single transaction or a transaction bundle in a designated future block. The providers for preconfs can be the block proposers or the block builders for Ethereum L1 preconfs, or the sequencer for L2 preconfs. Preconfs for L1 with block builders as commitment providers are live via mev-commit today.

A more advanced type of commitment are synchronous commitments, in which several providers for different chains participate. A synchronous commitment is only given if all participants agree to commit. The arrangement enables atomic execution across chains without any one provider trusting the others, because failure to comply triggers slashing for every deviator. Synchronous commitments extend the design space for the participating rollup ecosystems and unlock massive value by enhancing UX, increasing liquidity, and by enabling mainnet interoperability and efficient cross-chain operations.

The mev-commit network has been developed by Primev following early discussions with the Ethereum Foundation and Flashbots at Devcon Bogotá 2022. It was designed to support fast, private, and credibly neutral coordination across diverse actors in the MEV supply chain by enabling commitment games. The protocol operates through a dedicated coordination layer, the mev-commit chain, which is used to record and settle commitments in near real-time. All components of the mev-commit network are open-source (all code is available on GitHub). The protocol is chain-agnostic and supports any application that benefits from strong inclusion or execution guarantees, including bridges, searchers, rollups, and intent-based systems.

The mev-commit protocol enables commitment games as defined above. In its current deployment, it focuses on preconfirmations with Ethereum block builders as commitment providers. Bidders can specify in their bids which transactions, or transaction bundles, they want to have included at a certain block height. Furthermore, bidders can specify whether they only want inclusion guarantees, or whether they additionally want a commitment for the transactions executing without reverting. These commitments are posted to the mev-commit chain, an EVM chain with ~200ms block times (and as low as 5ms on testnet), optimized for high-throughput commitment settlement. Whenever a block builder builds the corresponding L1 block, this builder’s commitments will be checked for fulfillment and rewards or penalties will be paid between the bidders and this block builder. Commitments offer cryptographically secure privacy and their timeliness is incentivized via a decay mechanism.

We provide a high-level overview of the protocol below. For more detailed information, we refer to the mev-commit whitepaper [3] and the mev-commit documentation [4]. See Figure 1 below (taken from the whitepaper [3]) for a depiction of the basic interaction between the involved actors.

Privacy is essential to prevent strategic manipulation, frontrunning, or value leakage in MEV markets. To this end, mev-commit is designed with end-to-end privacy as a first-class feature. The protocol enforces three privacy guarantees:

1. Bid Privacy: Only intended providers can read the contents of a bid.

2. Commitment Privacy: Commitment contents remain hidden until explicitly settled.

3. Recipient Anonymity: Observers cannot determine which providers received a bid until it is revealed.

To achieve this, mev-commit introduces two cryptographic primitives:

• Anonymous Receiver-Updatable Broadcast Encryption (ARUBE) enables bidders to send encrypted bids to a dynamic set of providers without revealing the recipient list. This prevents attackers from deducing intent or frontrunning based on bid flow. It also provides recipient anonymity.

• Dual-Phase Commitments (DPCOM) allow both the bidder and provider to open a commitment, ensuring mutual accountability while maintaining confidentiality until settlement. If a provider fails to reveal a fulfilled commitment, the bidder can do so unilaterally.

These primitives combine to deliver a protocol that supports private coordination at scale, an essential requirement for use cases like private searcher flows, cross-chain bridging, and latency-sensitive applications. More information about privacy can be found in an article dedicated to it [5].

A key challenge in preconfirmation protocols is ensuring that providers issue commitments before the corresponding transaction is confirmed on-chain. Without safeguards, a provider could opportunistically wait to see a transaction included and then issue a "preconfirmation" retroactively, undermining its credibility. This issue is known as the preconfirmation fair exchange problem [6, 7].

mev-commit addresses this with a time-decay bidding mechanism. Each bid specifies a value function $$\beta(t)$$ that decreases linearly over time, starting from the moment the bid is issued and decaying to zero by a fixed expiry time. This creates strong cryptoeconomic pressure for providers to commit as early as possible, in particular before the relevant block is built. Late commitments receive reduced or no payment, making opportunistic behavior economically unviable.

The decay is enforced via timestamps recorded on the mev-commit chain, and validated using a timing tolerance parameter, which accounts for message latency. This ensures that honest providers with reasonable network conditions can participate without penalty, while discouraging timestamp manipulation.

By embedding decay directly into bid valuation, mev-commit solves the fair exchange problem in a decentralized setting, enabling credible commitments without trusted third parties.

The opt-in mechanism in mev-commit allows Ethereum validators to assert to only proposing blocks built by registered mev-commit builders. Validators signal their participation by registering their BLS public key either via a direct opt-in or via restaking through Eigenlayer or Symbiotic. In a rollup setting, this role can be played by sequencers as well as validators.

Opt-in plays a critical role in strengthening the reliability and enforceability of commitments: Users and bidders can place greater trust in commitments whenever the current validator is opted in, as the risk of non-inclusion due to proposing a block from a builder outside of mev-commit is mitigated. Furthermore, validators benefit economically from opting in by enabling higher-value blocks from builders who have received mev-commit bids. This increases builder competitiveness in mev-boost auctions and indirectly boosts proposer rewards.

mev-commit has seen substantial adoption across Ethereum’s consensus and execution layers. As of writing, more than 10,000 validators have opted into the protocol, signaling willingness to support blocks constructed by mev-commit registered builders.

Integration is also widespread across the builder and relay ecosystem. A majority of active block builders and relays support mev-commit, ensuring that a large portion of Ethereum’s blockspace can be coordinated through the protocol. This high level of builder participation enables bidders to source commitments from multiple providers, significantly increasing the probability that their desired execution paths are fulfilled.

This combination of builder coverage and validator opt-in ensures that mev-commit can offer near-instant execution assurances with strong economic backing.

Empirical data published on the Ethereum Research forum [8] document 807 execution preconfirmations across 415 Holesky blocks. Importantly, block value doubled: mev‑commit blocks averaged 0.0093 ETH versus 0.0044 ETH for vanilla counterparts. The figures confirm that commitment bids can significantly contribute to the overall block value. Mainnet growth continues developing, and can be tracked on the mev-commit Xplorer and this Dune Dashboard.

An Optimism sequencer wishing to issue commitments first stakes collateral on the mev-commit chain and registers its commitment key. It then launches a lightweight provider side‑car alongside the existing node. The side‑car’s duties are purely at the networking layer: receive bids labelled with the roll‑up’s chain‑ID, decrypt them, evaluate their fees, and, upon accepting, sign and send commitments. Importantly, no modification to the consensus or execution binaries is necessary. The sequencer only needs to ensure that commitments are only issued if fulfillment can be guaranteed, e.g., after already internally queuing the corresponding transactions for inclusion in the next block.

Note that mev-commit already works with multiple block builders on Ethereum L1. Using the same techniques as there, it is therefore also compatible with decentralized sequencers, where each node can run a provider node. The approach is thus future proof.

To enable synchronous commitments and atomic transaction bundles across different chains, a deeper integration is necessary. We here describe different alternatives how this could be achieved and their tradeoffs. The first approach via a dedicated synchronous commitment protocol provides the largest flexibility, but also requires all participants to engage in the protocol. As lightweight alternatives, we describe two related methods that still achieve some of the desired functionality with little extra effort from the commitment providers.

Synchronous Commitment Protocol. A bundle consisting of transactions for different chains requires coordination between the commitment providers of all these chains. For rollups, the providers are the respective sequencers and for Ethereum L1 the providers are all mev-commit registered block builders. A synchronous commitment should be issued if and only if all these providers agree on fulfilling their part of the transaction bundle (and if the relevant validator is opted in for L1 transactions), because only then atomic execution across all chains can be guaranteed.

Agreement either to include all transactions or none can be reached among the providers via a consensus protocol, either executed directly among themselves, or leveraging the mev-commit chain. Once consensus is reached, the corresponding commitment can be stored on the mev-commit chain. As for regular commitments, in case the synchronous commitment gets violated, the deviating providers get slashed and the corresponding bidder gets compensated.

Atomic Cross‑Chain Execution via State‑Lock Commitments. If some of the involved providers are not willing or able to participate in the consensus protocol described above, but instead can issue state-lock commitments, most of the functionality can still be achieved. This method has been suggested by AJ Park in the Flashbots research forum [9]. The basic idea is as follows: Bidders can obtain a state-lock commitment, granting the bidder exclusive access to a specified state (e.g., a specific Uniswap pool) for the next block. That is, after such a commitment is issued, transactions touching that state from everybody except the commitment’s bidder must be rejected, and transactions from the bidder must be included.

Given such state-lock commitments, bidders can coordinate atomic cross-chain bundles themselves as follows: First, bid for state-lock commitments for the relevant states on all involved chains. If such commitments are obtained from all chains, submit the corresponding transactions to all the chains. In that case, execution on all chains is guaranteed. If no state-lock commitment can be obtained on at least one chain in a timely manner, do not submit any of the involved transactions and potentially retry for the next block (or consider the bundle failed).

This method ensures that either all transactions in the bundle get executed on all chains, or none of them get executed (or the bidder gets compensated by a provider violating a commitment).

Atomic Cross‑Chain Execution via Cancellable Commitments. If state-lock commitments are not supported by all providers, a similar approach can be used via cancellable commitments: These are commitments that can be cancelled up to some deadline and if cancelled, the corresponding transactions must not be included. This method can even be employed if not all providers support cancellations. Consider, e.g., the scenario where cancellations are supported on L1, but not on some rollup and a user wants atomic execution across this rollup and L1. In that case, the user can first try to obtain a (cancellable) commitment from all providers on L1. If all these commitments are obtained, try to get a commitment on the rollup. If this is also obtained, all transactions are guaranteed to get executed (or the the bidder gets compensated). If no commitment can be obtained on the rollup, cancel the L1 commitments. In that case, no transaction will get executed, again ensuring atomic execution.

Bridging between Ethereum L1 and OP Stack rollups, or between different OP Stack chains, typically suffers from latency up to multiple days. The bottleneck is the time it takes for the origin chain transaction depositing funds into the bridge to reach confirmation or finality before the destination can release the funds. Sequencer Commitment Games (SCGs) remove this bottleneck by letting the origin chain’s block producer (L1 builder or L2 sequencer) issue an economically enforceable preconfirmation via the mev-commit network.

This source-chain preconfirmation cryptoeconomically guarantees that the bridge-initiating transaction will be included in a future block. Once a bridge operator or liquidity provider holds this guarantee, they can release funds instantly on the destination chain, because either (a) the origin transaction will appear as committed, or (b) the provider will be compensated from the violator’s collateral if it does not.

With this approach, instant bridging is possible in all directions between Ethereum L1 and different L2s. The pattern is the same for all directions:

• L1 → L2: L1 block builder commits to include the deposit transaction; destination L2 credits instantly.

• L2 → L1: L2 sequencer commits to include the withdrawal initiation; destination L1 credits instantly.

• Cross-rollup: Source rollup sequencer commits; destination rollup credits instantly.

See Figure 2 below for an overview of the general flow for an instant bridging transaction.

In today’s deployment, mev-commit preconfirmations are available for Ethereum L1 transactions, so instant bridging is live for L1-origin flows. Extending the same guarantees to L2-origin transfers requires integrating L2 sequencers as registered commitment providers.

The integration of Sequencer Commitment Games (SCGs) into OP Stack rollups provides a suite of technical, user-facing, and ecosystem-level advantages that address persistent challenges in scalability, interoperability, and liquidity.

By enabling low-latency preconfirmations, SCGs allow sequencers to issue economically credible inclusion or execution guarantees within as little as 5 ms. For traders, DeFi protocols, and wallets, SCGs enable instant trade confirmations, with execution guaranteed or compensated if deviations occur. This reduces transaction settlement uncertainty to near real-time and brings Layer 2 UX close to Web2 responsiveness.

As an additional benefit, mev-commit enables searchers to submit transaction bundles in the form of bids to rollups, a feature currently not supported natively by most rollups. In addition to simply submitting transaction bundles, which mev-searchers commonly do when interacting with L1 block builders, bundles via SCGs are economically protected: once committed to, their inclusion is backed by collateral, and any deviation triggers slashing. This means searchers get the guarantee that all transactions in the bundle get included in the same order as they appeared in the bundle and no other transactions will get in between, enabling complex searcher strategies.

Through integration with the mev-commit network and its pool of opted-in PBS actors on Ethereum mainnet, OP Stack rollups gain native interoperability with other rollups and Ethereum L1 without introducing new trust assumptions. Commitments can be coordinated with other sequencers and L1 block builders to enable seamless cross-domain workflows. Synchronous commitments can be used to enable atomic multi-chain execution. This increases capital efficiency by enabling sophisticated strategies such as atomic cross-chain arbitrage. As a related benefit, SCGs facilitate instant bridging between rollups and between L1 and L2, bypassing multi-minute to multi-day bridging delays. The resulting UX is trust-minimized, removing reliance on external bridge operators. All guarantees are enforced by the same cryptoeconomic security model as the underlying commitments.

The advantages are summarized in the table below, comparing the status quo with what can be achieved via SCGs:

Aspect                       | Status Quo             | With SCGs
-----------------------------|------------------------|-----------------
Transaction confirmation     | Up to 2 seconds        | As fast as 5 ms
L1 <-> Rollup Bridge Latency | Minutes to days        | Sub-seconds
MEV Bundle Support           | No                     | Yes
Validator/Relay Engagement   | No                     | Yes
Sequencer Modifications      | Required for upgrades  | None
UX Parity with Web2          | No                     | Much closer

A neutral, chain-abstracted commitment layer such as mev-commit creates new structural advantages for the OP Stack ecosystem, which are likely to increase TVL:

• Better liquidity routing: Commitments make liquidity across L1 and L2s more fungible by reducing settlement times, drastically reducing bridging delays, and enabling atomic execution, allowing to allocate capital where it earns the highest yield in real time.

• Support for cross-chain dApps: Developers can build applications that rely on fast, guaranteed execution across multiple rollups and L1, unlocking use cases that were previously infeasible due to settlement risk or latency.

• Shared revenue layer: Relays, sequencers, and validators participating in the commitment layer can share in new revenue streams generated by preconfirmation fees and cross-domain transaction flows.

The OP Stack is well-positioned to adopt SCGs early: its modular architecture, shared vision of the “Superchain,” and centralized-sequencer deployment model provide a straightforward integration path without protocol-level changes. Furthermore, new OP Stack rollups can onboard smoothly: once a sequencer integrates SCGs, it can immediately interoperate with Ethereum mainnet and other OP Stack chains through the shared commitment layer, without building bespoke infrastructure. To integrate, no customization of the sequencer logic is required and mev-commit is compatible with all kinds of sequencers, including those running inside a TEE.

Sequencer Commitment Games (SCGs) offer a practical, high-impact path to improving the UX, interoperability, and capital efficiency of the OP Stack ecosystem, without requiring any core protocol changes. By building on the proven mev-commit network, SCGs bring economically enforceable inclusion and execution guarantees to OP Stack sequencers, enabling low-latency preconfirmations, synchronous cross-chain coordination, and instant bridging between L1 and L2 as well as across rollups.

From a user perspective, SCGs deliver near-instant trade confirmations, trust-minimized bridging, and a Web2-like transaction experience. For searchers and liquidity providers, they unlock new mev strategies on L2s with cryptoeconomic protection against bundle breakage. They integrate seamlessly with Ethereum mainnet’s opted-in PBS actors, enabling native interoperability between rollups and L1.

At the ecosystem level, a neutral, chain-abstracted commitment layer can unify liquidity routing, support cross-chain dApps, and create shared revenue streams for participating actors, likely increasing TVL. The OP Stack is uniquely positioned to lead this effort: its modular design, Superchain vision, and sequencer architecture make adoption straightforward, allowing new rollups to plug into a shared commitment infrastructure from day one.

Key highlights include:

• No protocol-level changes: Integration happens via a side-car; sequencer logic stays untouched.

• Major UX gains: Instant confirmations and bridging, cross-rollup composability.

• Interoperability upside: Trust-minimized coordination with Ethereum L1 and other L2s.

• OP ecosystem leadership: Opportunity to set the standard for cross-chain coordination.

The research and development of SCGs for OP Stack rollups is at a pivotal stage. To realize this vision, collaboration across sequencer operators, rollup teams, wallets, and dApp developers is essential. We encourage the OP Stack community, core contributors, governance participants, and ecosystem partners to engage with the design process, test integrations, and provide feedback. With early adoption, the OP ecosystem can pioneer a new era of trust-minimized, chain-coordinated infrastructure and set a precedent for the broader Ethereum rollup landscape.

If supported by Optimism, Primev intents to continue working on the integration, in particular:

• Integration with Superchain sequencers, enabling synchronous commitments across the entire OP Stack ecosystem.

• Native wallet and provider tooling support, allowing users to request and verify commitments directly from their transaction interface.

• DAO onboarding of PBS actors, aligning governance with the economic incentives of block builders, relays, and sequencers to maximize adoption.

Acknowledgements. This research was supported by a grant from Optimism Foundation.

[1] Primev. Commitment Games and Where to Find Them, Mirror, 11 Jun 2024. https://mirror.xyz/preconf.eth/3SXageNPpcQMz8L2YfW5ipvthNl-InPzaPAfb3LxwqA

[2] Primev. Preconfirmations: The Fulfillment-Delivery Paradigm, Mirror, 7 Feb 2024. https://mirror.xyz/preconf.eth/sgcuSbd1jgaRXj9odSJW-_OlWIg6jcDREw1hUJnXtgI

[3] Primev. mev‑commit Whitepaper, v 1.0, 25 Nov 2024. https://whitepaper.mev-commit.xyz/

[4] Primev. mev-commit documentation. https://docs.primev.xyz/

[5] Primev. Pioneering End-to-End Privacy for Intra-Block Commitments, Mirror, 18 Sep 2024. https://mirror.xyz/preconf.eth/iz2J0uTXHhl8DiAkG-VLLwvCp-8qcc_Z7A8_4rU0A3g

[6] Justin Drake. Based preconfirmations. Ethereum Research, 2023. https://ethresear.ch/t/based-preconfirmations/.

[7] Yaonam. Solutions to the preconf fair exchange problem. Ethereum Research, 2024. https://ethresear.ch/t/solutions-to-the-preconf-fair-exchange-problem/.

[8] Evan Kim. Preconfirmation Bidding Increased Block Values on Holesky, ethresear.ch, 2024. https://ethresear.ch/t/preconfirmation-bidding-increased-block-values-on-holesky/

[9] AJ Park. Atomic Cross‑Rollup Arbitrage via Coordinator and State‑Lock Auctions, The Flashbots Collective, 12 Jun 2025. https://collective.flashbots.net/t/atomic-cross-rollup-arbitrage-via-coordinator-and-state-lock-auctions/