ERC-4337 is an Ethereum Improvement Proposal authored in 2021 by Vitalik Buterin, Yoav Weiss, Dror Tirosh, and four additional contributors. It introduces account abstraction — allowing smart contract wallets to behave like native Ethereum accounts — without any changes to the Ethereum consensus protocol.

The key insight: rather than modifying Ethereum’s transaction pool (mempool) or validator rules, ERC-4337 introduces a separate, higher-layer infrastructure: a new transaction type called UserOperation, a new mempool for UserOperations, Bundlers who package UserOperations into standard Ethereum transactions, and a singleton EntryPoint smart contract that handles all validation and execution.

> Full EIP: eips.ethereum.org/EIPS/eip-4337

---

Publication and Context

Before ERC-4337, Ethereum had two account types:

1. Externally Owned Accounts (EOA): Controlled by private keys. Standard user accounts (MetaMask). Simple but inflexible — whoever holds the key controls the account; no multi-sig, no recovery.
2. Contract Accounts: Smart contracts with programmable logic but cannot initiate transactions on their own; must be triggered by an EOA.

This created a painful UX limitation: every Ethereum user must hold ETH for gas fees and secure a private key with no recovery mechanism. Loss of the private key means permanent loss of funds.

Account abstraction had been a long-standing Ethereum research goal. Multiple previous EIPs (EIP-86, EIP-2938) proposed consensus-level changes but were never adopted due to complexity. ERC-4337 solves account abstraction without touching consensus — making it deployable immediately on any EVM chain.

ERC-4337 deployed to Ethereum mainnet on March 1, 2023 (with EntryPoint v0.6). EIP-7702 (2024) later added a complementary simpler mechanism for EOA-to-contract upgrades.

---

Architecture

UserOperations

A UserOperation (UserOp) is a new pseudo-transaction type (not a standard Ethereum transaction):

UserOperation {
  sender: address           // Smart contract wallet address
  nonce: uint256            // Replay protection
  initCode: bytes           // Wallet factory call (if first op)
  callData: bytes           // What the wallet should do
  callGasLimit: uint256
  verificationGasLimit: uint256
  preVerificationGas: uint256
  maxFeePerGas: uint256
  maxPriorityFeePerGas: uint256
  paymasterAndData: bytes   // Optional paymaster contract + data
  signature: bytes          // Arbitrary signature (wallet-defined)
}

Key difference from standard tx: The signature field can be anything — multi-sig, BLS, passkey, time-lock, hardware key — because signature validation logic is in the wallet contract, not the protocol.

Bundlers

Bundlers are infrastructure nodes that:

1. Monitor the UserOperation mempool (separate from the standard tx mempool)
2. Simulate UserOps locally to verify they are valid (and that they will pay)
3. Bundle multiple UserOps into a single standard Ethereum transaction calling the EntryPoint
4. Submit the bundle to Ethereum (paying ETH gas themselves; recouped from wallet gas payments)

Anyone can run a Bundler; they earn gas fee margins for their service. Bundlers must stake ETH in the EntryPoint to participate (preventing spam).

EntryPoint

The EntryPoint is a singleton smart contract (deployed at a fixed address on every EVM chain):

1. Receives bundled UserOps from Bundlers
2. Validates each UserOp: calls the wallet contract’s validateUserOp() function
3. Executes each UserOp: calls the wallet contract to perform the actual action
4. Charges gas fees from the wallet (or Paymaster)
5. Pays the Bundler for gas costs

The EntryPoint has been audited by OpenZeppelin, Trail of Bits, and others.

Paymasters

Paymasters are optional smart contracts that can sponsor gas fees:

paymasterAndData = [paymaster_address][paymaster_data]

If a UserOp includes a Paymaster:

• The Paymaster’s validatePaymasterUserOp() is called instead of charging the wallet
• The Paymaster can pay gas in ERC-20 tokens (e.g., USDC for gas)
• The Paymaster can provide fully sponsored gasless transactions (dApp pays for the user’s gas)

This enables: users who hold only USDC (no ETH) can execute Ethereum transactions; dApps can offer free trial transactions.

---

Enabled Features

Feature	How ERC-4337 enables it
Social recovery	Wallet contract has guardian logic; if key is lost, guardians can authorize new key
Multi-signature	Wallet’s validateUserOp() requires M-of-N signatures
Gas sponsorship	Paymaster pays gas instead of user
ERC-20 gas	Paymaster accepts token payment; pays ETH gas
Session keys	Wallet allows sub-keys with limited permissions and expiry
Passkey auth	Wallet validates WebAuthn/passkey signatures (P-256 curve)
Batch transactions	Single UserOp calls multiple contracts
Automatic payments	Wallet logic can auto-approve recurring charges

---

Reality Check

ERC-4337 is deployed and functional. As of late 2024:

• 30+ million UserOperations processed on EVM chains
• Major wallet providers (Coinbase Smart Wallet, Safe, Biconomy, Alchemy’s LightAccount) use ERC-4337
• EntryPoint v0.7 deployed (2024) with improved gas efficiency

Limitations:

• Gas overhead: UserOps cost more gas than standard EOA transactions (~20–50k additional gas per UserOp) due to EntryPoint overhead
• Bundler centralization: In practice, a small set of Bundlers (Biconomy, Pimlico, Stackup) process most UserOps; Bundler centralization is a liveness risk
• UX still complex: Despite account abstraction, wallet onboarding remains complex for normie users
• EIP-7702 competition: A simpler 2024 EIP allows EOAs to temporarily behave like smart contracts without full ERC-4337 infrastructure, potentially reducing AA adoption for simpler use cases

---

Legacy

ERC-4337 is the most impactful Ethereum infrastructure upgrade that required no Ethereum consensus change. It enables the next generation of user-friendly crypto wallets. Paymasters for gasless transactions and session keys for dApp-specific permissions are already deployed and used by millions of users.

---

Related Terms

• Account Abstraction
• Smart Contract Wallet
• Gas
• Paymaster
• Bundler

---

Research

• Buterin, V., Weiss, Y., Tirosh, D., Nacson, S., Forshtat, A., Gazso, K., & Hess, T. (2021). ERC-4337: Account Abstraction Using Alt Mempool. EIPs.ethereum.org.

— The primary EIP; full specification of UserOperation format, EntryPoint contract interface, Bundler validation rules, and Paymaster design.

• Daian, P., Goldfeder, S., Kell, T., Li, Y., Zhao, X., Bentov, I., Breidenbach, L., & Juels, A. (2020). Flash Boys 2.0: Frontrunning in Decentralized Exchanges, MEV, and Consensus Instability. IEEE S&P 2020.

— MEV context relevant to Bundlers: Bundlers face MEV opportunities within UserOperation bundles; the alt-mempool design partially mitigates standard MEV but introduces Bundler-level MEV.

• Wood, G. (2014). Ethereum: A Secure Decentralised Generalised Transaction Ledger (Yellow Paper). ethereum.github.io.

— Ethereum EVM and transaction model specification; ERC-4337 is designed specifically to complement and extend this without modifying it.