When a trade gets sandwiched or front-run, the money extracted from it usually goes to whoever controls the ordering of transactions in that block – and that is not always the bot that spotted the opportunity in the first place. Ethereum’s own developer documentation says that for competitive strategies such as DEX arbitrage, searchers may have to pay 90% or more of their total MEV revenue back to the validator as gas fees, simply to guarantee their transaction lands first, according to ethereum.org.
What MEV is, mechanically
MEV stands for maximal extractable value: the profit that whoever decides the order, inclusion or exclusion of transactions in a block can capture, according to ethereum.org’s developer documentation and Eco.com’s support article on the topic. The term began as “miner extractable value” because miners controlled block construction under proof-of-work; after Ethereum moved to proof-of-stake at the Merge in September 2022, the same role passed to validators, and the term became “maximal extractable value,” per ethereum.org and Eco.com. Eco.com’s account extends this further, describing the broader research community as having generalized the term to cover any block-ordering actor on any chain – a cross-chain framing that is Eco.com’s own and is not made in the ethereum.org documentation reviewed for this page.
On Ethereum today, that job is split across three roles under a system called proposer-builder separation, per Eco.com. Searchers scan the mempool – the holding area where submitted transactions wait before confirmation – for profitable sequences, and submit them as bundles to builders. Builders assemble full candidate blocks out of many searchers’ bundles plus ordinary transactions, competing to offer the most total value. Validators then pick the highest-paying block from the builder market and sign it, without seeing its contents in advance, according to Eco.com. The validator is not choosing which trade to allow through; it is choosing which block pays it the most.
The three extraction patterns
Eco.com and ethereum.org describe three overlapping patterns. Front-running: a searcher spots a pending trade that will move a price and places an identical-direction trade ahead of it with a higher fee, landing first and capturing the price move before the original transaction executes, per Eco.com. Back-running: a searcher takes the other side immediately after a known transaction creates an imbalance – the clearest case is a lending-protocol liquidation, where searchers race to be first to call the liquidation function and collect the reward, described in similar terms by ethereum.org and Eco.com. Sandwiching combines both: one trade placed immediately before a victim’s transaction and one immediately after it, extracting the price impact the victim’s own trade caused – described independently by ethereum.org, EigenPhi’s classroom documentation, Arkham’s guide, MEV Blocker’s documentation and Eco.com.
A worked example – and why it is illustrative, not a receipt
MEV Blocker’s documentation walks through a hypothetical trade to demonstrate the mechanics. In its example, a trader wants to buy 2,000 LINEA for roughly 1 ETH but sets a 10% slippage tolerance, meaning she will accept paying up to 1.1 ETH. A searcher spots the pending order, buys 2,000 LINEA for 1 ETH just ahead of it, which pushes the price to her tolerance ceiling; her trade then clears at 1.1 ETH, and the searcher immediately sells the same LINEA back for 1.1 ETH, netting a gross profit of 0.1 ETH before gas costs, per MEV Blocker’s documentation.
EigenPhi’s classroom documentation runs a similar hypothetical with different numbers. A trader selling 1,000 WETH for USDC gets sandwiched by a searcher who sells 650 WETH ahead of the trade at an exchange rate of 1 WETH for 2,834 USDC, pushing the rate down to 1 WETH for 2,821 USDC before the victim’s own trade executes. The searcher’s round trip nets 2.9 WETH in gross revenue; EigenPhi’s documentation assumes 1.2 WETH in gas and miner tips, leaving a profit of 1.7 WETH.
Both walkthroughs come from documentation pages built to teach the arithmetic, not from a cited on-chain transaction – neither EigenPhi’s nor MEV Blocker’s page names a transaction hash, block number or date for the trade it describes. Treat the dollar and ETH figures in both as worked arithmetic, not as records of something that happened.
Two figures in this evidence set are different in kind: ethereum.org’s documentation presents them as specific real-world transactions rather than teaching arithmetic, though the underlying transaction records sit behind external links this page did not independently review. Ethereum.org describes an arbitrage transaction in which a searcher turned 1,000 ETH into 1,045 ETH by exploiting a price gap on the same ETH/DAI pair between Uniswap and Sushiswap, executed atomically in a single transaction, and links out to the transaction itself. Separately, ethereum.org describes a case in which a searcher spent $7 million buying every available CryptoPunk at floor price in one MEV-driven transaction, working with an MEV provider to keep the purchase concealed until it executed, and links to further reporting on it.
Where the money actually lands
The natural assumption is that the bot which finds the opportunity keeps the reward. Ethereum.org’s own documentation complicates that for the most competitive strategies: because many searchers can spot the same DEX arbitrage at once, the only way to guarantee inclusion is to outbid every other searcher on gas price, and a rational searcher will keep raising that bid up to the value of the opportunity itself. Ethereum.org states that searchers may end up paying 90% or more of their total MEV revenue in gas fees to the validator, precisely because so many bots chase the identical trade. The validator collects this without needing to run any MEV-detection infrastructure of its own – under proposer-builder separation, it simply accepts the highest bid a builder offers, per Eco.com.
Research from the University of Copenhagen and eToro Denmark, published as a conference talk proposal on arXiv, adds a structural detail: at the time of writing, the paper states, roughly 90% of active Ethereum validator nodes were running a standard implementation of the MEV-Boost software that receives builder bids, which means the bidding mechanism the paper describes was not a marginal arrangement even at that point. The same paper collected bid data submitted by builders through relays over a dated window, from 15 September 2022 through 31 January 2023, to study multi-block MEV strategies. Separately, in an “at the time of writing” figure the paper does not explicitly tie to that collection window, it states that MEV opportunities have generated 182,456 ETH in total; its authors describe the findings as preliminary and non-conclusive, and the document itself is a talk proposal rather than a peer-reviewed study.
How much MEV costs users, in total
No single agreed figure exists for the total value extracted from crypto users through MEV. Arkham’s article, published 8 July 2026, cites EigenPhi’s data to state that arbitrage MEV generated $3.37 million in profit over a 30-day period in September 2025. Eco.com’s support article, separately, says EigenPhi’s own dashboard shows sandwich and arbitrage MEV producing eight-figure monthly totals on Ethereum, with a running total in the billions of dollars over the chain’s history, and notes that Flashbots’ MEV-Explore tracks the same activity with a different methodology and reaches only “comparable” orders of magnitude – not the same number. Neither tracker’s underlying dashboard was reviewed directly for this page, so none of these totals can be independently checked here.
What this page does not tell you
- It does not give a current, reconciled dollar total for how much MEV has extracted from crypto users. EigenPhi, Flashbots’ MEV-Explore and Arkham use different methodologies and, per Eco.com’s own account, only converge on “comparable” orders of magnitude, and none of the underlying dashboards were part of the evidence reviewed for this page.
- The EigenPhi and MEV Blocker sandwich walkthroughs used above to explain the mechanism are worked examples from documentation pages built to teach the arithmetic. Neither cites a transaction hash, block number or date, so they should not be read as records of specific trades that occurred.
- Arkham’s article, published 8 July 2026, states that enshrined proposer-builder separation is “now the standard” on Ethereum, and separately claims Solana has become a significant venue for MEV despite lacking a public mempool, relying instead on Jito bundles and private relays. Neither claim is corroborated by any other source reviewed for this page.
- The arXiv paper’s figure of 182,456 ETH in MEV opportunities is given as an “at the time of writing” total that the paper does not explicitly link to its separate 15 September 2022 to 31 January 2023 bid-collection window; the paper is a preliminary conference talk proposal rather than a peer-reviewed or continuously updated study, and neither figure describes current MEV volumes.
- None of the sources reviewed here quantify how much of the value extracted specifically through sandwich attacks, as opposed to arbitrage or liquidations, ends up with validators rather than with the searchers who found it.
Sources
Every fact above is attributed to one of these reports. Where they disagree, the article says so.
