Ethereum Block Builders: The Hidden Engine of MEV
Block builders control which transactions make it on Ethereum, and just 2 of them build ~86% of all blocks. Here's how they work and who dominates.
Key takeaways
- Block builders and validators are separate roles. Builders construct blocks, while validators only sign and propose them
- MEV-Boost is the current implementation of PBS: off-chain middleware connecting builders, relays, and validators – used by over 90% of the network
- Two builders currently control most of Ethereum's block production, creating real censorship risk even with a decentralized validator set
- Regular users are directly affected: sandwich attacks, front-running, and transaction censorship are downstream consequences of how the builder market is structured
Ethereum block builders are specialized entities that assemble transaction blocks on behalf of validators, competing in real-time auctions to have their block selected and added to the chain. Rather than building blocks themselves, validators delegate this job to builders via MEV-Boost – a system now used by over 90% of the Ethereum network.
Understanding block builders means understanding who actually controls what gets included in each Ethereum block and why that matters far beyond MEV.
What Is an Ethereum Block Builder?
| Quick answer: An Ethereum block builder is an entity that constructs full execution blocks, including selecting, ordering, and bundling transactions, and submits them to validators through a competitive auction. |
After The Merge in 2022, Ethereum moved from proof-of-work to proof-of-stake. Validators replaced miners as block proposers, but a new question emerged: who should build the block? Validators are randomly selected and may lack the technical sophistication to construct maximally profitable blocks. Block builders filled that gap.
A builder's job is to take the raw ingredient of pending transactions, assemble the most valuable block possible, and submit it as a bid. The validator simply picks the highest bid and signs it.
Why Block Builders Exist: The MEV Problem
| In short: Block builders exist because transaction ordering on Ethereum is worth a lot of money, and extracting that value requires specialized expertise that most validators do not have. |
MEV (Maximal Extractable Value) is the additional profit that can be captured by choosing which transactions to include in a block and in what order. Common MEV strategies include:
- Arbitrage: Exploiting price differences across DEXs before they close
- Liquidations: Triggering undercollateralized loans for a fee
- Sandwich attacks: Front-running and back-running a user's trade to extract value from the price movement
In early 2025, approximately $289 million was extracted through sandwich attacks alone. Over the full year of 2023, roughly $3.2 billion in MEV was extracted on Ethereum.
If validators extract MEV themselves, the ones with the most sophisticated infrastructure accumulate disproportionate rewards, pushing out solo stakers and centralizing the validator set.
The solution Ethereum settled on: separate the builder from the proposer. Let experts build blocks and compete for the right to propose. Validators collect the auction revenue without needing to run MEV infrastructure themselves.
How Ethereum Block Builders Work
| In short: A block builder's core job is to construct the single most valuable block it can within a 12-second slot window, then win the auction to have it included. The process moves through four stages. |
Collecting transactions from the mempool
Builders monitor the public mempool – Ethereum's global waiting room, where all pending transactions sit before inclusion. Every wallet submission, swap, and DeFi interaction passes through here before landing in a block.
But builders don't rely only on the public mempool. They also receive transactions through private order flow channels, such as direct pipelines from wallets, DeFi protocols, and trading bots that want their transactions included quickly and without exposure to front-running.
As of 2026, over 50% of Ethereum gas originates from private or protected transaction channels, meaning the majority of high-value activity bypasses the public mempool entirely.
Searching for MEV opportunities
Once builders have their pool of transactions, they analyze the current state of DeFi protocols to identify profitable orderings. This is often done in collaboration with searchers – specialized bots that scan for MEV opportunities and submit pre-packaged transaction bundles to builders.
Searchers always route their transactions privately, either through known private channels or via direct communication with builders. Builders that have exclusive relationships with high-quality searchers consistently outperform those relying on the public mempool alone.
Constructing the most profitable block
With transactions and bundles in hand, the builder solves an optimization problem: given limited block space, limited gas, and possible transaction conflicts – what ordering maximizes total value?
This includes:
- Sequencing arbitrage bundles at the right moment in the block
- Including liquidations when collateral thresholds are breached
- Ensuring conflicting transactions don't cancel each other out
- Deciding which low-fee transactions are worth including vs. skipping
The result is a fully assembled block with a specific bid – the amount the builder is willing to pay the validator in exchange for having this block proposed.
Submitting bids to validators
Builders don't submit blocks directly to validators. Instead, they route through relays – trusted intermediaries that aggregate bids from multiple builders and serve the most profitable block header to validators.
The relay holds the block payload in escrow. The validator sees only the bid amount and a commitment. Once the validator signs the header, the relay releases the full payload.
This design prevents validators from stealing the block's MEV while also preventing builders from reneging on their payment. An auction cycle lasts approximately 12 seconds, after which the winning block is selected, and a new cycle begins.
BytebyByte's take: MEV-Boost was designed to democratize MEV rewards. Validators of all sizes get auction revenue without needing specialized infrastructure. And it worked for validators. But the competition shifted entirely to the builder layer, where exclusive order flow agreements created a market structure that no protocol rule was designed to prevent. The result is a duopoly that controls the majority of Ethereum block production through business development. PBS solved the validator centralization problem by creating a new centralization problem one layer up. That's a tradeoff Ethereum's designers didn't fully anticipate.
How Proposer-Builder Separation (PBS) Changed Ethereum
In short: PBS transformed Ethereum block production by splitting a single role, the validator who both built and proposed blocks, into two separate, specialized actors:
|
This separation is what made block builders a permanent fixture of the Ethereum ecosystem.
Before PBS, Ethereum validators (and before that, miners) handled both jobs. They selected transactions from the mempool and appended the block themselves. This worked at small scale, but as DeFi grew, the MEV embedded in each block became large enough that only sophisticated, well-capitalized operators could reliably extract it.
Ethereum researchers proposed PBS as a structural fix: let validators stay simple proposers, and let specialized builders compete on block construction.
MEV-Boost, built by Flashbots, implemented PBS off-chain starting in 2022. It's middleware that validators run alongside their beacon node, connecting them to a marketplace of builders via relays. Within months of launch, adoption exceeded 60% of validators; today it exceeds 90%.
The practical outcome:
- Solo stakers can now compete for MEV revenue without running sophisticated bots
- Block construction professionalized rapidly into a competitive market
- Relay operators became an unexpected trust chokepoint in the system
PBS via MEV-Boost is explicitly described as a temporary solution – a "proto-PBS", while Ethereum's protocol engineers work on enshrining it natively. That effort is ePBS, covered in the final section.
Ethereum Block Builders vs Searchers vs Validators
These three roles are often confused because they interact tightly, but each has a distinct function in the block production pipeline
Actor | Primary Role | Makes money from |
| Searcher | Finds MEV opportunities; submits transaction bundles | Profit from arbitrage, liquidations, sandwich strategies |
| Block Builder | Assembles full blocks from transactions and bundles | Bid margin after paying the validator |
| Validator (Proposer) | Selects the winning block and signs it | Auction revenue from builder bids |
Searchers are automated bots. They constantly scan mempool data, DeFi protocol states, and price feeds to spot profitable opportunities. When they find one, they construct a bundle – a precise sequence of transactions that must execute in a specific order – and submit it to one or more builders.
Block builders aggregate input from searchers, private order flow providers, and the public mempool. Their job is to turn raw transaction data into the most profitable possible block. Some builders also run their own internal searchers, which gives them consistent, proprietary order flow.
Validators are mostly passive in this pipeline. They register with relays, receive block bids, and select the highest-paying one. Their staking rewards increase substantially. MEV-Boost is reported to raise validator rewards by 20–50% compared to local block building.
The three actors form a vertical stack. And increasingly, the top builders are vertically integrated, running their own searchers, controlling private order flow, and effectively capturing value at multiple layers of the stack.
Leading Ethereum Block Builders
| In short: The Ethereum block-building market is highly concentrated. As of early 2025, two builders, Beaverbuild and Titan Builder, controlled approximately 86% of Ethereum mainnet block production. |
Current market share breakdown (Rated Network, cumulative data):
Builder | Market Share |
| Beaverbuild | ~35–50% |
| Titan Builder | ~23–37% |
| Rsync | ~10–13% |
| Flashbots/BuilderNet | ~9% |
| Builder 0x69 | ~8% |
| Others | < 5% |
Using the Herfindahl-Hirschman Index (HHI), a standard market concentration metric, the Ethereum builder market scores approximately 3,892, far above the 1,800 threshold the U.S. Department of Justice defines as "highly concentrated".
Why did this duopoly form?
The key driver is exclusive order flow agreements. Deals where a wallet, DeFi protocol, or trading platform sends its transaction flow to only one builder. These agreements directly translate into higher block value, which translates into winning more auctions.
- Titan Builder secured an exclusive deal with Banana Gun, the dominant Telegram trading bot during the memecoin boom on Ethereum
- Beaverbuild partnered with CoW Swap and other private flow providers
Rsync lost its competitive footing after failing to secure comparable exclusive deals. Its market share declined sharply after 2024 Q1, despite having a competitive technical infrastructure.
Builder market share is determined more by business development than technical excellence.
What Ethereum Block Builders Mean for Regular Users
For most Ethereum users, block builders are invisible, but they directly affect the outcome of every transaction.
When you submit a swap on Uniswap or buy a token on a DEX, your transaction enters the mempool before it's included in a block. Searchers monitoring the mempool can see your trade and construct bundles that execute around it:
- Front-run: buy before your trade pushes the price up
- Back-run: sell after your trade moves the price
- Sandwich: both – bracketing your transaction to extract the price impact
A 2025 Flashbots study found that 1.2% of all DEX trades on Ethereum are sandwiched, with an average loss of 0.41% of trade value per attack. For a user making $10,000 in monthly DEX trades, that represents roughly $590 in hidden losses per year.
What's Next: ePBS and the Road to Decentralized Block Building
| In short: The current block-building architecture with off-chain relays serving as trusted intermediaries was always designed as a temporary measure. The long-term solution is ePBS (Enshrined Proposer-Builder Separation), tracked as EIP-7732. |
What ePBS does
Under ePBS, the proposer-builder auction is built directly into Ethereum's consensus protocol. Builders would assemble blocks and cryptographically seal their contents; proposers would select the highest-paying block without being able to see or tamper with what's inside. Transactions are revealed only after the block is finalized.
The relay – currently an off-chain, trusted intermediary – is removed as a chokepoint. The protocol itself manages the trustless exchange of a block payload for payment.
Vitalik Buterin acknowledged the limits of this change directly: "ePBS only prevents builder centralization from spilling over into the staking layer, but the issue of builder centralization itself remains." The duopoly among builders is an economic problem, and ePBS doesn't directly fix that.
Where things stand in 2026
ePBS is scheduled as a headline feature of the Glamsterdam upgrade, Ethereum's next major hard fork. As of April 2026, ePBS implementation is described by the Ethereum Foundation as "trickier than anticipated," with devnet testing actively underway.
Current estimates target mainnet activation around late Q3 2026, though timeline slippage is possible.
Also on the roadmap
- FOCIL (Fork-choice Enforced Inclusion Lists) – gives validators the power to force inclusion of transactions that builders might otherwise ignore; targeted for the subsequent Hegotá upgrade
- Flashbots BuilderNet – a near-term collaborative block-building network that distributes builder work across multiple participants before ePBS is live
Neither ePBS nor FOCIL fully solves builder concentration – but together, they reduce the systemic risks that concentration creates.
Sources and Further Reading
- Flashbots – "MEV-Boost: Proposer-Builder Separation for Ethereum" https://github.com/flashbots/mev-boost
- Flashbots Docs – "MEV-Boost Overview" https://docs.flashbots.net/flashbots-mev-boost/introduction
- arXiv 2311.09083 – "Structural Advantages for Integrated Builders in MEV-Boost" https://arxiv.org/pdf/2311.09083
- arXiv 2509.16052 – "How Exclusive Are Ethereum Transactions? Evidence from Non-Winning Blocks" https://arxiv.org/pdf/2509.16052
- arXiv 2506.04940 – "Becoming Immutable: How Ethereum Is Made" https://arxiv.org/pdf/2506.04940
- ethereum.org – "Glamsterdam Upgrade" https://ethereum.org/roadmap/glamsterdam/
- Ethereum Foundation Blog – "Checkpoint #9: April 2026" https://blog.ethereum.org/2026/04/10/checkpoint-9
- Etherscan Info – "Unbundling the Current State of Proposer-Builder Separation" https://info.etherscan.com/unbundling-the-current-state-of-pbs/
FAQs About Ethereum Block Builders
Yes. Builders have full discretion over which transactions to include in their blocks. A builder might exclude a transaction if it's unprofitable to include, if it conflicts with a higher-value bundle, or if the builder has chosen to comply with external compliance requirements (such as OFAC sanctions screening).
