Intents Are Eating ERC-4337

TL;DR

• Userops (ERC-4337) are slow, complex, and high-cost.
• Intents are fast, simple, and low-cost.
• Intents can be chain agnostic (ERC-7683). Ethereum’s horizontal scaling strategy almost guarantees that intents will eat userops.

Introduction

Account abstraction is the end game for the mainstream adoption of Ethereum-compatible blockchains. In 2023, ERC-4337 was introduced as the transaction infrastructure to enable smart accounts in a permissionless and decentralized manner. We believe this infrastructure will remain foundational to Ethereum’s account model remaining trustless and censorship resistant. However, this infrastructure’s complexity, cost, and latency limit the widespread adoption of smart accounts. Furthermore, Ethereum’s liquidity fragmentation across many L2s necessitates a transaction infrastructure that is dynamic, lightweight, and chain agnostic.

Intents and standards like ERC-7683 are eating userops. They present a number of advantages: cost efficiency, speed, developer simplicity, and the ability to be easily extended to support cross-chain transactions. In our 2025 predictions, we wrote that ERC-4337 would be met with new competition. The competition is intents, and the adoption is already taking off, with many account abstraction players making the switch.

It’s important to highlight that intents and userops are not mutually exclusive. They are complementary. The complexity and overhead of ERC-4337 come from its open and permissionless properties. We believe ERC-4337 is foundational to ensuring that Ethereum’s account model remains trustless. But for speed and best-in-class UX, intents are taking over.

What Are Userops?

ERC-4337 introduces infrastructure at the application layer of Ethereum to enable account abstraction. That is, enabling smart contract accounts (herein “Smart Accounts”) to transact without relying on an Externally Owned Account (EOA). The introduction of Smart Accounts makes the user account programmable and unlocks many new quality-of-life features for end users (e.g., gas abstraction and batching).

ERC-4337 infrastructure enables Smart Accounts by introducing User Operations (userops), which are pseudo-transactions that contain transaction details to be executed on behalf of the user. Userops are sent to a dedicated mempool, bundled by specialized actors called bundlers, and submitted to the blockchain via a global entrypoint contract.

Key Features of Userops:

• Batching Transactions: Users can perform multiple actions (e.g., token approvals and swaps) in one transaction.
• Flexible Gas Payments: Gas fees can be paid in ERC-20 tokens or by third parties (i.e., paymasters).
• Isolated Validation Phase: The separation or userop validation and execution, and the introduction of opcode restrictions in the validation phase, allows for an open mempool that is DOS resistant.

The scope of userops is focused on single-chain interactions, enhancing functionality and user experience within a specific blockchain like Ethereum. Systems that extend userops to multiple chains, such as Biconomy’s MEE, do so by bundling userops and sequentially executing them through traditional bridging routes. These routes are packed into a single signature envelope signed by the user.

What Are Intents?

Intents are user-defined goals like swapping tokens between two chains or staking assets across networks. A network of fillers (or solvers) competes to execute these intents efficiently across multiple blockchains.

ERC-7683 standardizes cross-chain intents, allowing users to specify desired outcomes across blockchains without worrying about the underlying execution routing. Intents abstract away the complexities of the underlying chain or protocol delivering the required transactions to achieve the desired outcome.

Key Features of Intents:

• Cross-Chain Interoperability: Intents enable seamless actions across different blockchains, eliminating the need for users to manage multiple wallets or bridges.
• Simplified User Experience: Users specify what they want (e.g., “Swap Token A on Chain X for Token B on Chain Y”), and fillers handle how to achieve it.
• Standardized Execution: ERC-7683 provides a unified framework for expressing and fulfilling intents, making it easier for developers and fillers to participate.

The scope of intents is, by definition, broad. It includes same- and cross-chain interactions, addressing the challenges of operating in a fragmented, multi-chain ecosystem.

Why Intents Are Eating Userops

Account abstraction is a key transition for Ethereum as we move from a niche technology into mainstream adoption. It introduces many quality-of-life improvements that are necessary for the vast majority of the population to engage with blockchain-based and self-sovereign systems.

ERC-4337 brought the first wave of early adopters and a significant increase in the tooling and services available to developers building on smart accounts. However, it did not bring about a sizable rise in Smart Account users. One driver for this was the friction in migrating from an existing EOA to a Smart Account — EIP-7702 resolves this and will be included in Pectra. The second is the complexity and overhead ERC-4337 introduces, creating friction for developer adoption and increasing the cost of transacting (gas).

The core drivers for intents eating userops:

1. Enhanced cross-chain liquidity and efficiency
2. Faster execution
3. Cheaper gas costs
4. Simplified developer and user experience

Enhanced Cross-Chain Liquidity and Efficiency

The proliferation of layer 2s, especially appchains, means users increasingly need to interact across chains. Intents provide a unified way to handle these interactions, reducing friction and improving accessibility. Userops, while powerful, are limited to single-chain environments and do not address cross-chain needs without adding extra layers of complexity. Furthermore, paymasters restrict locked funds to a specific gas operation on a single chain, limiting inventory utilization.

Intents, however, provide solutions that enable a shared network of solvers who can compete to fulfill user requests and absorb the added complexity. This competition can lead to better pricing for complex cross-chain routes (e.g., those that include swaps) and faster execution across chains, resulting in a more efficient use of liquidity across the ecosystem. The flexibility in solver actions also guarantees a much higher utilization rate of stored inventory.

Faster Execution

The ERC-4337 process has multiple stages, especially those involving paymasters, introducing significant RPC latency. A sponsored userop has multiple round trips between the ERC-4337 infrastructure and the wallet, resulting in meaningful RPC latency. On the other hand, intents utilize a network of fillers who compete to execute the user’s request as quickly and efficiently as possible. Since much of the work (e.g., finding the best execution path or liquidity source) happens off-chain, intents avoid the bottlenecks of onchain processing that userops require. Fillers can respond to real-time conditions — such as selecting the fastest DEX or intent-bridge — ensuring quicker execution than the rigid userops pipeline.

Cheaper Gas Costs

Intents are significantly cheaper regarding gas fees because they avoid the expensive infrastructure required by ERC-4337. ERC-4337 relies on a multi-layered system to process user operations: paymasters, bundlers, and a global entrypoint contract. Each component adds overhead that increases gas costs. The entrypoint contract is a coordination hub where every userop is validated, gas payments are handled, and execution occurs. This multi-step process requires additional onchain computations, driving up fees. Paymasters, while useful for gas abstraction, introduce extra onchain operations to sponsor or manage gas payments. Bundlers aggregate userops and submit them to the entry point, but this aggregation and submission process also consumes gas, especially as each userop is processed individually.

Intents streamline the process by offloading much of the execution logic to fillers (solvers) who compete off-chain to fulfill the user’s request. This reduces the number of onchain interactions. A user simply signs an intent specifying their desired outcome (e.g., purchase an NFT on Polygon using ETH on Base), and fillers handle the execution, often optimizing paths or batching multiple intents into a single transaction. By minimizing onchain steps and leveraging off-chain competition, intents significantly lower gas costs compared to the heavy infrastructure of userops.

Simplified Developer and User Experience

Intents allow developers and users to focus on outcomes rather than processes. For example, a user can express an intent to swap tokens across chains without understanding bridges or gas fees on different networks. This can be translated to a significant improvement in the developer experience. For example, when a developer uses Omni Account, the only input required from the developer is the destination chain and the desired outcome (e.g., an output token or DeFi action); our path algorithm and solver network do the rest.

Developers can build applications that leverage intents to offer seamless cross-chain functionality without integrating with multiple bridges or protocols individually. This reduces development overhead and fosters innovation, as developers can focus on building user-centric features rather than handling cross-chain complexities. Userops, while useful for single-chain applications, do not provide the same level of flexibility for multi-chain development. They also require a bunch of complex tooling — bundlers for gas estimation, complex simulations, etc — while intents are standard transactions without the need for these extra bells and whistles.

Complementary Nature of Userops and Intents

Userops and intents are not mutually exclusive and can complement each other. For Example, Omni Account is compatible with both intents and userops.

ERC-4337 provides a trustless and censorship-resistant method for Smart Accounts to transact. Although intents provide more than features parity with ERC-4337 bundlers and paymasters (as explained above), it is incredibly difficult to replicate the fully decentralized nature of ERC-4337’s open mempool. By supporting this infrastructure, developers can ensure that in all cases, their users are resistant to censorship, allowing for true self-sovereign autonomy on the blockchain.

In summary, intents provide the performance, efficient, and abstracted transaction rails for Smart Accounts, whilst the ERC-4337 infra provides a trustless vital fallback mechanism.

Conclusion

Conceptually, intents and userops are very similar. However, the flexibility of intents and the relaxation of the constraint to provide censorship resistance enables a drastic simplification of intent-based infrastructure, unlocking speed, flexibility, cost-effectiveness, and a chain abstracted substitute to userops.

In many cases, applications interface with private bundlers, negating the trustless properties of ERC-4337 whilst adopting all the complexity. Intents will eat this order flow and ERC-4337 will fall back to its foundations — censorship-resistant transaction rails for smart accounts.