Most blockchains run a public waiting room. Transactions sit in the mempool, readable by anyone, until a block includes them, and that transparency is what makes the network verifiable. It is also what lets specialized bots study your pending swap and act on it before it confirms. MEV is the umbrella term for the value those ordering games can extract.
The scale stopped being theoretical years ago: Flashbots, the research organization that first measured it, counted at least $675 million of MEV extracted on Ethereum between January 2020 and the network's switch to proof of stake in September 2022.
What is MEV, exactly?
Maximal extractable value is the most that whoever orders transactions can gain by including, excluding or reordering them inside a block. Not all of it costs you anything: much of MEV is arbitrage that realigns prices between venues after ordinary trades. The part aimed at users is different: front-running, where a bot places its own trade ahead of your visible one, and the sandwich attack, which wraps your swap from both sides and takes the difference.
A useful mental model: MEV is rent on visibility. Anything a stranger can predict about your transaction before it settles, its pair, its size, its worst acceptable price, is information that can be priced against you. Every effective protection either reduces what strangers can see or binds what they can do with it.
Where did MEV come from?
The term began as miner extractable value in Flash Boys 2.0, a 2019 research paper that documented bots openly bidding against each other to reorder trades on early DEXes. After Ethereum moved to proof of stake, the name generalized to maximal extractable value, because validators, block builders and searcher bots now share the game. The mechanics survived the rename intact.
Today extraction is a professionalized pipeline: searchers hunt opportunities, builders assemble blocks and validators auction off the right to order them. That is also why waiting for the problem to fade is not a plan. Ordering rights are worth real money, and someone will always hold them.
How does a sandwich attack work?
- 1
The bot spots your pending swap
Your transaction waits in the public mempool with its pair, size and slippage tolerance readable by anyone who looks.
- 2
It buys ahead of you
The bot places the same swap just before yours in the block, pushing the pool's price up toward your tolerance limit.
- 3
Your swap fills at the worse price
You receive close to the minimum your tolerance allows. The gap between that and the quoted rate is what the bot came for.
- 4
It sells right behind you
In the same block, the bot unwinds at the price your swap just pushed up, and the loop closes.
The pattern is industrial. EigenPhi data covering November 2024 to October 2025 counts more than 95,000 sandwich attacks on Ethereum, running 60,000 to 90,000 per month, with roughly 70% of them traced to a single operator. Almost 40% hit stablecoin pools, exactly the pairs traders assume are too quiet to attack.
Stablecoin pairs attract the attacks for two reasons. Their prices are so steady that any tolerance beyond a few hundredths of a percent is free room for a bot to move in, and traders grant that room casually because the pair feels safe. The feeling is exactly what gets monetized.
How much does MEV cost traders?
Cointelegraph Research, working on the same EigenPhi data set, puts sandwich losses at about $60 million a year on Ethereum. The direction of travel is the encouraging part: monthly extraction fell from about $10 million in late 2024 to about $2.5 million by October 2025, even as monthly DEX volume climbed past $100 billion, and the researchers attribute part of that decline to traders adopting MEV protection.
Averages hide the tail. Most sandwich fills skim a fraction of a percent, but the worst cases are brutal: in one widely reported incident from March 2025, a $220,000 stablecoin swap on Ethereum lost about $215,000 to a single sandwich attack. The cost of MEV arrives as a distribution, and your worst day sits somewhere in it.
How does MEV protection work on 1inch?
A sandwich needs to see your swap before it settles, so the effective protection is structural: take the swap out of the public mempool entirely. In the default intent-based mode on 1inch you sign an order stating what you give and the minimum you receive. Resolvers, professional market participants, compete to fill it and submit the on-chain transaction themselves. Your order never sits in the public queue, so there is nothing to front-run and nothing to sandwich, and cross-chain swaps carry the same protection.
The order's terms are enforced on-chain: a fill below the rate you signed cannot settle, no matter what the mempool is doing around it. That turns MEV protection from a promise into a property of the transaction, which is the difference that matters when the attacker is faster than you.
- Visibility
- A public-mempool swap is readable by every bot before it confirms. An intent-based order goes to resolvers, not the public queue.
- Who submits
- You broadcast a public transaction yourself. A resolver submits the intent-based fill and pays its gas.
- Rate at settlement
- A public swap can fill anywhere down to your tolerance. An intent-based fill settles at the signed rate or better, or not at all.
- Cost of failure
- A reverted public swap still pays gas. An unfilled order costs nothing.
If you do submit public transactions, in the Pro interface or elsewhere, the practical hygiene is the same everywhere: keep slippage tolerance tight, prefer deep pools for large swaps and treat any fill that lands exactly on your limit as a signal worth reading. Structural protection is better than hygiene, but hygiene still narrows the target.
What should you read next?
Slippage tolerance is the dial sandwich bots exploit, and routing is what keeps rates honest. Read what slippage is and what a DEX aggregator does, or browse all Learn guides.
Frequently asked questions
No. Arbitrage that realigns prices across venues is MEV too, and it mostly costs no one at the moment it happens. The forms worth defending against are the ones priced off your pending transaction: front-running and sandwich attacks.
No, it is part of how intent-based orders settle rather than an add-on. The quote you accept already reflects it, alongside competitive rates, aggregated across 300+ sources.
A sandwich requires a visible pending transaction, and an intent-based order never becomes one. The minimum you signed is enforced on-chain, so a fill below it cannot settle regardless of what bots do.
Yes. The ordering game exists wherever transactions queue publicly; Ethereum's mempool is just the most studied arena, and layer-2 sequencers and other chains have their own variants. Cross-chain swaps on 1inch carry the same intent-based protection as single-chain ones.
Swap where bots cannot see you
Intent-based swaps on 1inch settle privately, with MEV protection included by default.
