What is MEV in crypto?

MEV is the value that block proposers and arbitrage bots can capture by controlling the order of transactions in a block. Because blocks are produced sequentially and the proposer chooses which transactions to include and in what order, profitable opportunities exist to insert, reorder, or omit user transactions. The most common MEV strategy is the sandwich attack — front-running a victim swap to push price up, then back-running to sell at the inflated price.

What does MEV cost the average user?

MEV extracted from Ethereum users alone exceeds $1B per year. The most visible cost shows up as worse swap execution (sandwich-attacked trades) — typically 0.1-2% extra slippage. AMM swaps on public mempools are most exposed; private order flow (Flashbots Protect, MEV-Share) and intent-based DEXes (CoW Swap, UniswapX) significantly reduce user MEV exposure.

How do AI agents protect against MEV?

Three primary tools: (1) submit transactions through MEV-protected RPCs (Flashbots Protect, MEV Blocker) which use private mempools. (2) Use intent-based DEX aggregators (CoW Swap, UniswapX) that batch and protect orders. (3) Set tight slippage tolerances and use private order flow on Solana (Jito for blockspace auctions). Hive's tool surface includes MEV-aware swap routing recommendations as part of analyze_coin and trade-execution context.

No. MEV from arbitrage and liquidations is value-additive — it keeps DEX prices aligned across venues and removes underwater positions promptly. The harmful subset is user-extractive MEV (sandwich attacks, front-running). The crypto industry's response has been to formalize MEV markets (Flashbots, Jito) so the unavoidable extraction goes through transparent auctions rather than dark forest tactics.

What is MEV? | Maximal Extractable Value Definition for Crypto Agents

MEV — Maximal Extractable Value — is the profit a block proposer or searcher can extract by including, reordering, or excluding transactions in a block.

The acronym originally stood for "Miner Extractable Value" (when Ethereum was proof-of-work and miners ordered blocks), and was renamed after the Merge to reflect that any block proposer — validators, builders, searchers — can extract this value.

Common MEV strategies:

Arbitrage — capturing price differences across DEXes by ordering trades in a single block. Generally value-additive.
Liquidations — racing to liquidate underwater lending positions for the liquidation bonus. Generally value-additive.
Sandwich attacks — front-running a victim's AMM swap to push the price up, then back-running to sell at the inflated price. The victim sees worse execution; the searcher pockets the spread. Extractive.
Front-running — inserting an agent's transaction ahead of a known profitable trade to capture the opportunity first. Extractive.
Time-bandit reorgs (rare on Ethereum, more common on lower-finality chains) — rewriting the chain to capture MEV from a previous block.

Why agents must account for MEV

An AI agent that recommends a token swap without MEV awareness is recommending a trade that may be worse than the user expected — sometimes 1-2% worse on a sandwiched swap. Three mitigations: (1) use MEV-protected RPCs for execution, (2) route through intent-based aggregators (CoW Swap, UniswapX), (3) set tight slippage tolerances.

Hive's tool surface includes MEV-aware execution context in analyze_coin and trade-recommendation tools. CoinGecko MCP and Moralis Cortex do not surface MEV-protection guidance.

Why agents must account for MEV

Related terms