MEV-Boost Relay: The Heart of Ethereum PBS
90%+ of Ethereum blocks pass through MEV-Boost relays, yet most validators don't know how they work. Here's what relays actually do and who runs them.
Key takeaways
- A MEV-Boost relay is a middleware layer in Ethereum's Proposer-Builder Separation (PBS) system that connects block builders and validators without exposing block contents prematurely.
- Relays exist because vanilla Proof-of-Stake gives validators full control over transaction ordering, creating MEV centralization incentives that undermine network fairness.
- Relay infrastructure is currently centralized. A handful of operators route the majority of MEV-Boost blocks, with measurable censorship implications.
- The long-term solution is Enshrined PBS (ePBS), which moves the builder auction into Ethereum's consensus layer, but it remains an open research problem as of 2025–2026.
A MEV-Boost relay plays the role of an auctioneer in Ethereum's block production pipeline. It receives competing blocks from builders, surfaces the highest bid to validators, and holds the full block in escrow until the validator commits. Without it, the builder market cannot function safely.
Understanding how relays work is key to understanding where Ethereum's block production actually happens and why it matters for decentralization, censorship resistance, and staking rewards.
What Is a MEV-Boost Relay?
| Quick answer: A MEV-Boost relay is a piece of middleware infrastructure that acts as a trusted bridge between block builders and Ethereum validators in the PBS pipeline. It receives complete blocks from competing builders, validates them, and presents only the block header, along with the bid value, to the validator for selection. |
The relay never lets the validator see the full block content before they commit to it. This is the mechanism that prevents validators from extracting the builder's MEV opportunities for themselves.
Once the validator signs the header, the relay releases the full block for broadcast to the network.
MEV-Boost itself is the open-source software, developed by Flashbots, that validators run as a sidecar alongside their consensus client. The relay is the external service that MEV-Boost connects to. A single validator can register with multiple relays simultaneously, increasing their chances of receiving the highest-value block at any given slot.
Why Does Ethereum Need MEV-Boost Relays?
| Quick answer: MEV-Boost relays exist because the original Proof-of-Stake design gave validators too much power over block construction and too little incentive to use it fairly. Relays, as part of the PBS architecture, are the practical solution that Ethereum deployed to address this structural problem. |
The limitations of vanilla Proof-of-Stake
In a standard PoS setup, the validator is selected to propose block builds that block themselves. They choose which transactions to include, in what order, and at what fee threshold. This means the same entity controls both who gets in and in what sequence – a combination that creates significant MEV extraction opportunities.
Sophisticated validators can run searcher bots to front-run pending transactions, sandwich DeFi trades, or perform arbitrage between DEXes. Less sophisticated validators cannot compete with these strategies, leading to an uneven playing field where larger, well-resourced operators earn disproportionately more.
Left unchecked, this dynamic pushes solo stakers out and concentrates stake in a few professional operations.
The rise of Proposer-Builder Separation (PBS)
PBS is the architectural fix: split the job of building a block from the job of proposing it.
Role | Who does it | What they do |
| Builder | Specialized firms | Assemble maximally profitable blocks |
| Proposer (Validator) | Any validator | Select and propose the best block offered |
This division benefits all parties:
- Builders compete on block quality, driving up bids
- Validators earn more without running complex MEV strategies, estimates suggest over 60% higher staking rewards versus local block building
- The network gets more efficient, and competitive block construction
PBS was not part of Ethereum's original Merge design. MEV-Boost was Flashbots' out-of-protocol implementation – a community-built stopgap deployed while native, protocol-level PBS remained under research.
The role of relays in PBS
Without a relay, PBS faces a fundamental trust problem:
- If a validator sees the full block before committing → they can steal the builder's MEV
- If a validator cannot see the block before committing → they have no way to verify the bid
The relay resolves this by acting as an escrow and validator:
- Builder submits a full block to the relay
- Relay validates the block and confirms the bid value
- Relay sends only the block header to the validator, not the full contents
- Validator signs the header as a commitment
- Relay releases the full block. The validator's signature makes any deviation slashable
This enforces honest behavior through economic penalties rather than cryptographic guarantees.
How Does a MEV-Boost Relay Work? (Step-by-Step)
| Quick answer: A MEV-Boost relay works by running a sealed-bid auction. Builders submit full blocks, the relay surfaces only the highest bid to validators, and releases the full block only after the validator commits, ensuring neither side can cheat the other. |
The full pipeline runs through four stages, with the relay sitting at the center of the process.
Step 1: Builders create candidate blocks
Searchers continuously scan the public mempool and on-chain state for profitable MEV opportunities, such as arbitrage between DEXes, liquidations, or sandwich positions. They package these into transaction bundles and send them to block builders.
Block builders collect bundles from multiple searchers, combine them with regular mempool transactions, and simulate the full block to find the combination that maximizes total revenue. A single slot can see dozens of competing builder submissions.
Step 2: Builders send bids to relays
Once a builder has assembled their optimal block, they submit it to one or more relays, along with a bid indicating how much they will pay the validator if their block is selected. The relay receives the full block, validates it for correctness (checking signatures, balances, state transitions), and stores it in escrow.
Builders typically submit to multiple relays to maximize their chances of winning the auction. Each relay independently tracks the highest bid it has received and makes that information available to validators.
Step 3: Relays send the best bid to validators
When a validator's slot approaches, MEV-Boost queries all connected relays for their best current bid. Each relay responds with a block header and the corresponding bid value, but not the full block. MEV-Boost selects the highest bid across all relays and returns that header to the validator's consensus client.
The validator signs the header. This signature is the commitment. MEV-Boost sends the signed header back to the winning relay, which treats this as a binding agreement: the validator has promised to propose this specific block.
Step 4: Block publication
Upon receiving the signed header, the relay releases the full block payload to the validator. The validator's consensus client broadcasts the complete block to the Ethereum network for attestation and inclusion. The builder's bid is paid to the validator's fee recipient address, typically via a direct ETH transfer at the end of the block.
If the relay fails to deliver the payload after receiving the signed header, the validator misses their proposal slot. They cannot fall back to a locally built block at that point without risking a double-proposal. This is one of the key trust assumptions in the current system.
Author's note: What the relay really is
The more you look at how MEV-Boost relays operate, the more you realize they are opinionated gatekeepers. A relay decides which builders can submit blocks. It decides which transactions it will validate. And it can simply refuse to propagate blocks containing transactions it dislikes. None of this is visible to the average user or even most validators. The relay is the hidden hand in Ethereum's block production pipeline: technically optional, practically unavoidable, and quietly shaping which transactions get included and which ones don't. The shift to ePBS is Ethereum's attempt to stop outsourcing a critical governance function to a handful of private companies operating under their own compliance policies.
>> Learn more: MEV Supply Chain Explained: Searchers to Validators
The Relay Landscape: Who Are the Major Players?
As of mid-2026, only around eight active relays operate on the Ethereum mainnet, and the top five together route over 90% of all MEV-Boost blocks. The market is small, concentrated, and split along one critical axis: whether or not the relay censors OFAC-sanctioned transactions.
Relay | Operator | OFAC-Compliant | Notes |
| Flashbots | Flashbots | Yes | Pioneer relay; historically dominant market share |
| bloXroute Max Profit | bloXroute | No | Prioritizes profitability exclusively |
| bloXroute Regulated | bloXroute | Yes | Compliance-focused version |
| Ultra Sound | ultrasound.money | No | Non-censoring; strong community trust |
| Aestus | Aestus | No | Neutral, non-censoring |
| Agnostic Gnosis | Gnosis | No | Censorship-free; run by Gnosis DAO |
| Titan | Titan Builder | No | Neutral; growing market share |
The concentration problem is significant. Only eight or so active relays exist on mainnet, and the top five together account for over 90% of relayed blocks. That level of concentration in infrastructure that routes the majority of Ethereum's block production is a well-recognized centralization risk.
Censoring vs. non-censoring relays is the most operationally important distinction for validators today. Censoring relays filter blocks that include transactions interacting with OFAC-sanctioned addresses – most notably addresses associated with Tornado Cash after the U.S. Treasury sanctions in 2022. Non-censoring relays propagate all valid transactions regardless of regulatory origin.
When evaluating relays, validators typically weigh three factors:
- Profitability: how competitive the relay's bids are, based on the builder pool it connects to
- Censorship posture: whether the relay filters OFAC-sanctioned transactions
- Reliability: uptime, response latency, and track record of delivering payloads on time
Tools like relayscan.io and mevwatch.info provide real-time data on relay performance and censorship metrics.
Risks and Criticisms of the Current Relay Model
| In short: The current relay model has three core structural problems: it centralizes critical infrastructure in a few hands, enables transaction censorship by private operators, and creates trust dependencies that no cryptographic guarantee backs up. |
Centralization risk
~90% of all Ethereum blocks now flow through the MEV-Boost pipeline, making the relay layer one of the most critical, and least decentralized, components in the network.
Relay-side concentration:
- Only ~8 active relays exist on mainnet
- The top 5 relays route over 90% of all MEV-Boost blocks
- A disruption or coordinated action among just a few operators could have an outsized impact on block production
Builder-side concentration amplifies the problem:
- ~90% of Ethereum blocks are produced by just 4 block builders
- The leading builder alone commands close to 50% market share
- Vertical integration, controlling both builder and relay infrastructure, creates latency and orderflow advantages that smaller participants cannot match
Censorship and OFAC compliance
OFAC-compliant relays refuse to propagate blocks containing transactions from sanctioned addresses. After the U.S. Treasury sanctioned Tornado Cash in August 2022, this became a live, measurable issue.
What happened:
- At peak, over 70% of MEV-Boost blocks were OFAC-compliant
- Sanctioned-address transactions were systematically excluded from the majority of block flow
The practical effect is not an outright ban. A transaction can still land if a non-censoring validator builds locally or routes through a neutral relay. But the consequences are real:
- Longer inclusion times
- Worse UX for affected users
- A two-tier system where inclusion depends on which relay a validator happens to use
Credible neutrality, the principle that Ethereum includes any valid, fee-paying transaction, erodes when private US-based companies set the filtering policies for the majority of block production.
Trust assumptions & relay incentive misalignment
Relays sit at the most sensitive point in the pipeline, yet operate on pure reputation — no cryptographic enforcement, no on-chain accountability.
Who trusts what:
Party | Trust placed in relay |
| Builders | Relay won't front-run or copy their block contents |
| Validators | Relay will deliver the payload after the header is signed |
Neither trust assumption is backed by protocol rules. Both rely entirely on the relay's reputation and fear of reputational damage.
The incentive problem:
- Builders earn MEV → clear financial incentive to compete
- Validators earn staking rewards → clear financial incentive to participate
- Relays earn nothing directly → operate as a public good or strategic investment
This misalignment creates a sustainability risk: if the relay operation becomes costly, operators may exit, concentrating the market further.
Timing games add another layer. Validators who delay their block request to the last millisecond can capture higher MEV bids, but validators with higher-latency infrastructure lose this game systematically, creating an arms race that favors well-capitalized professional operators.
The Future of MEV-Boost Relay: ePBS, FOCIL and Beyond
| In short: The long-term goal is to eliminate trusted relays by moving PBS into Ethereum's consensus protocol – a proposal called Enshrined PBS (ePBS). Until those ships are built, relays remain necessary infrastructure, but the roadmap is clear. |
Three proposals are shaping where this goes: ePBS tackles the relay trust problem at the root, FOCIL addresses censorship independently, and MEV Burn changes the economic incentive structure altogether.
Enshrined PBS (ePBS)
The primary proposal. ePBS moves the builder auction directly into Ethereum's consensus rules.
- Builders bid on-chain
- The protocol enforces payment and payload delivery
- No trusted intermediary needed at any step
- Listed as an Ethereum Foundation 2026 priority
The tradeoff – the free option problem:
Under some ePBS designs, a builder can win the auction and then withhold the payload if market conditions shift against them in the final seconds. The outcome:
Action | Consequence for the builder |
| Deliver payload | Pay the bid, block goes through |
| Withhold payload | Liveness failure, but no direct financial penalty |
This gives builders a risk-free short-dated option on block rewards – a structural asymmetry that active research is still working to solve.
FOCIL (Fork-Choice Enforced Inclusion Lists)
FOCIL attacks the censorship problem from a different angle by making censorship technically unenforceable at the builder level.
- Randomly selected validators submit mandatory inclusion lists for each slot
- Builders must include those transactions for their block to be valid under fork-choice rules
- A relay or builder that filters OFAC-sanctioned transactions cannot override a FOCIL-listed transaction
Result: transaction inclusion becomes a protocol-level guarantee, not a relay policy decision.
MEV Burn
A complementary proposal that changes where builder bids go.
- Instead of being paid to the proposing validator → bids are burned
- MEV rewards are smoothed across all ETH holders via reduced supply
- Reduces the economic arms race that drives builder and relay centralization
None of these are live on mainnet. MEV-Boost relays remain the operating reality for the foreseeable future, which is precisely why their dynamics matter now.
Should Validators Care About Which Relay They Use?
| Quick answer: Yes, relay selection affects three things: staking returns, censorship posture, and the overall health of Ethereum's block production infrastructure. |
- For profitability
Not all relays offer the same bid quality. Relays with access to a larger or more competitive builder pool tend to surface higher bids. Connecting to multiple relays is generally advisable, as MEV-Boost will always select the highest bid across all connected relays for a given slot. However, adding more relays also increases the risk of encountering a poorly performing relay that could cause a missed proposal.
- For censorship posture
Solo validators and smaller operators are the only participants in the ecosystem who can freely choose non-censoring relays without regulatory obligations. Centralized exchanges and custodial staking providers typically operate under compliance constraints that require OFAC adherence.
If you are a solo staker and value Ethereum's censorship resistance, routing blocks through neutral relays like Ultra Sound, Aestus, Agnostic Gnosis, or Titan has zero cost to your rewards and a positive impact on network neutrality.
- For network health
The relay market is already highly concentrated. Validators who consciously diversify their relay connections contribute to a more resilient block production pipeline, even if the marginal difference per validator is small.
Sources and Further Reading
- Flashbots – "MEV-Boost Introduction" https://docs.flashbots.net/flashbots-mev-boost/introduction
- Flashbots – "MEV-Boost GitHub Repository" https://github.com/flashbots/mev-boost
- Ethereum.org – "Maximal Extractable Value (MEV)" https://ethereum.org/en/developers/docs/mev/
- Umbra Research – "Ethereum's Supply Chain, Part 2: Towards ePBS" https://www.umbraresearch.xyz/writings/ethereum-supply-chain-part-2
- Etherscan Info –"Unbundling the Current State of Proposer-Builder Separation" https://info.etherscan.com/unbundling-the-current-state-of-pbs/
- Fenbushi Capital – "The Pursuit of Relay Incentivization" https://fenbushi.vc/2024/01/25/the-pursuit-of-relay-incentivization/
- MEV Watch – "Censoring Relay Share on Ethereum MEV-Boost" https://www.mevwatch.info/
- relayscan.io – "Real-Time Relay Performance Data" https://relayscan.io
- Ethereum Research Forum – "ePBS Design Discussions" https://ethresear.ch/tag/epbs
FAQs About MEV-Boost Relay
Technically, yes. A relay sees the full block contents and could theoretically extract or replicate its MEV opportunities before propagating it. This is why builder trust in relays is one of the most debated issues in the ecosystem. In practice, established relays like Flashbots have maintained reputational commitments to neutrality, but this protection is social, not cryptographic.
