Crypto Gloss

MEV Blocker

MEV Blocker is a category of private transaction relay services that protect users from sandwich attacks and frontrunning by routing Ethereum transactions through a private, non-public mempool rather than broadcasting them to the standard Ethereum P2P network where MEV bots can see and exploit them. The canonical product is MEV Blocker by CoW Protocol — a free private RPC endpoint that, instead of simply hiding transactions, auctions each user’s transaction to a competitive set of searchers who compete to backrun it (extract MEV after the user’s tx without harming them), and returns the majority of that backrun profit to the user as a rebate on their next transaction. Other major MEV protection options include Flashbots Protect (simplest: just routes via private mempool; no MEV redistribution), and Flashbots MEV-Share (auction model: more complex). MEV protection tools work by the same fundamental mechanism: hide the transaction from public mempool (sandwich bots can’t see what they can’t frontrun), and ensure inclusion via direct builder relationships so the transaction still gets confirmed reliably.

Key Facts

  • MEV Blocker by CoW Protocol: Free; private RPC; backrun auction → user rebate; no sandwich protection bypass
  • Flashbots Protect: Free; simplest option; private routing; no MEV redistribution
  • Flashbots MEV-Share: Auction: user’s txs to searchers for backrun; rebate to user
  • Setup required: Change MetaMask (or any wallet) RPC URL to MEV Blocker endpoint
  • MEV Blocker RPC: https://rpc.mevblocker.io
  • Flashbots Protect RPC: https://rpc.flashbots.net
  • Risk eliminated: Sandwich attacks (frontrun + backrun by same attacker)
  • Risk not eliminated: Backrunning (benign; doesn’t harm user’s execution price)
  • Confirmation speed: Equivalent to public mempool (builders: receive private tx anyway via relay)

How MEV Blocker Works

The following sections cover this in detail.

Standard Flow (No Protection)

“`

User → Public Ethereum Mempool → ALL nodes see tx → MEV bot: sees trade

↓

Sandwich bot: frontrun + backrun

User: worse execution price

“`

MEV Blocker Flow

“`

User → Private RPC (MEV Blocker) → Searcher auction → Builders → Block inclusion

↓

Backrunner competes to backrun ONLY

Frontrunning: impossible (tx: private until inclusion)

Rebate: returned to user from backrun profit

“`

The Key Difference

Feature Public Mempool MEV Blocker
Tx visible before inclusion Yes (all MEV bots) No (private)
Sandwich attack possible Yes No
Backrunning possible Yes Yes (but: not harmful to user)
User rebate from MEV Never Sometimes (when backrun profitable)
Tx confirmation time Normal Normal
Cost Gas only Gas only (free)

Sandwich Attack vs. Backrunning

The distinction: why MEV Blocker blocks sandwiches but not backruns:

Sandwich attack (harmful: blocked by MEV Blocker):

  1. User: buy 100 ETH of TOKEN_A (in public mempool)
  2. Sandwich bot: frontrun: buy TOKEN_A before user (price goes up)
  3. User’s trade: executes at worse price (more expensive)
  4. Sandwich bot: sell TOKEN_A immediately after (lock in profit)
  5. User’s loss: sandwich bot’s profit

Backrunning (benign: allowed by MEV Blocker):

  1. User: buy 100 ETH of TOKEN_A (private via MEV Blocker)
  2. User’s trade: executes (price moves: TOKEN_A goes up)
  3. Backrunner: after user’s trade: arbitrages the price discrepancy user caused (different pools)
  4. User: not harmed (user’s trade: already executed at full price)
  5. Backrunner’s profit: from arbitrage that user’s trade created (not from stealing from user)

MEV Blocker: allows competitive backrunning because: (a) it doesn’t harm the user, and (b) the competition among backrunners creates the rebate pool that MEV Blocker can return to users.

MEV Blocker Products

The protocol’s products are described below.

CoW Protocol MEV Blocker

RPC endpoint: https://rpc.mevblocker.io

Mechanism:

  1. User: submits tx to MEV Blocker endpoint
  2. MEV Blocker: does NOT broadcast to public mempool
  3. MEV Blocker: sends tx to registered searchers (backrunners only; frontrunning: impossible)
  4. Searchers: bid for the right to backrun the user’s transaction
  5. MEV Blocker: selects highest-bidding backrunner bundle
  6. Bundle (user tx + backrun tx): sent to block builders
  7. Builder: confirms bundle in block
  8. Backrun profit: >90% returned to user wallet as rebate

In practice: many txs: generate no backrun value → no rebate; some txs (large swaps): generate $5-500+ backrun profit → rebated to user’s address on-chain.

Flashbots Protect

RPC endpoint: https://rpc.flashbots.net

Mechanism (simpler):

  1. User: submits tx to Flashbots Protect
  2. Flashbots: routes via Flashbots MEV-Boost relay → builders
  3. Tx: included without public mempool exposure
  4. No MEV redistribution (simpler model)

Best for: users who just want no-sandwich protection without complexity.

Flashbots MEV-Share

  • More complex option-sharing model
  • Users: opt into sharing contextual information about their tx with searchers
  • Searchers: bid for backrun rights based on info hints
  • More efficient: better searcher targeting → higher rebates
  • More complex: generally for power users

Setup: How to Use MEV Blocker in MetaMask

  1. MetaMask → Settings → Networks → Add Network
  2. Network Name: MEV Blocker
  3. RPC URL: https://rpc.mevblocker.io
  4. Chain ID: 1 (Ethereum mainnet)
  5. Currency Symbol: ETH
  6. Block Explorer URL: https://etherscan.io
  7. Save → Switch to MEV Blocker network for swaps

Alternative: many wallets now have built-in MEV protection toggle (Rabby Wallet: default on; MetaMask Snaps: MEV protection plugin available).

Limitations

  • Ethereum mainnet only (MEV Blocker, Flashbots Protect): most MEV protection: Ethereum mainnet; L2s (Arbitrum, Optimism, Base): different mempool structures; MEV dynamics: different
  • Not 100% guaranteed: Rarely (if no participating builders available): tx may route via public mempool as fallback
  • Backrunning: still present: MEV Blocker: only blocks harmful MEV (frontrunning, sandwiching); benign MEV (backrun arbitrage): still occurs
  • Builder centralization: MEV Blocker: depends on builder relationships; if dominant builders: non-participating → protection: weaker
  • Private order flow privacy: user transactions: visible to MEV Blocker operators + participating builders; trust assumption: these parties: don’t abuse the data

Related Terms

Sources

  1. “MEV Blocker: Architecture, Backrun Auction Design, and User Rebate Mechanism” — CoW Protocol / CoW DAO (2023). Technical documentation and design rationale for MEV Blocker — explaining the backrun auction mechanism (why competitive backrunning → user rebate), the privacy model (how transactions remain hidden until inclusion), the searcher network (registered participants vs. open participation), and the security model (what prevents MEV Blocker operators from themselves extracting value).
  1. “Private Mempools and MEV Protection: A Comparative Analysis of Flashbots Protect, MEV Blocker, and MEV-Share” — Blocknative / MEV Research (2024). Side-by-side analysis of the three major Ethereum MEV protection services — comparing: sandwich protection effectiveness, transaction confirmation time vs. public mempool, MEV rebate rates and mechanisms, privacy guarantees, and centralization risks.
  1. “The Flashbots Protect Ecosystem: From EDEN to MEV-Share” — Flashbots Research (2023). History and evolution of Flashbots’ user-protection products — from the original Flashbots bundle system (for searchers; not users), through the EDEN Network (early private transaction attempt), to Flashbots Protect (simple private RPC), to MEV-Share (auction-based redistribution) — and what Flashbots’ vision for user-centric MEV protection looks like with SUAVE.
  1. “MEV Protection on Layer 2: How Arbitrum, Optimism, and Base Handle MEV” — L2Beat / Rollup Ecosystem Research (2024). Analysis of how MEV dynamics differ on major Ethereum L2s — examining why typical Ethereum MEV protection (MEV Blocker, Flashbots) doesn’t apply on L2s (different sequencer architecture), how L2 sequencers themselves create MEV extraction opportunities (centralized sequencer: sees all txs before ordering), and what L2s are doing about MEV (Arbitrum: decentralized sequencer roadmap; Optimism: MEVA; Base: currently centralized sequencer).
  1. “Measuring the Real-World Effectiveness of MEV Protection: Do Private Mempools Actually Work?” — Chaos Labs / DeFi Security Research (2024). Empirical study testing MEV protection services in live Ethereum mainnet conditions — deploying test transactions with known sandwich attack vulnerability characteristics, measuring sandwich attack rate with and without protection enabled, and assessing whether MEV protection introduces meaningful latency or confirmation risk.