In a traditional on-chain swap, a user submits a transaction that specifies exact routing parameters (path, pools, limits) and pays gas to execute it. The transaction hits the public mempool, which exposes it to reordering and sandwich attacks if the parameters are loose.
In intent-based trading, the user signs an order that expresses an outcome constraint, such as:
A third party then decides how to satisfy the intent and pays the execution costs. That third party is typically called a solver, filler, or resolver, depending on the protocol.
The difference is not semantic. It changes who controls routing, who pays gas, and where MEV can be extracted.
Intent systems generally have three layers:
The competition layer is where most tradeoffs live.
CoW Protocol is a canonical example of solver competition. User intents are grouped into batches, and solvers submit solutions for each batch. The winning solver is selected as the one that provides the most surplus to user orders, then settles the batch on-chain.
This design shifts MEV from “miners/validators extract from users” toward “solvers compete to deliver value to users,” but it also introduces a specialized execution class that must be trusted to behave within the protocol’s rules.
UniswapX is an intent-based protocol where fillers compete to execute user swaps. Swappers create orders defining auction parameters and price tolerance, and the auction mechanism varies by chain.
The important detail is that fillers can source liquidity from many places (AMMs, RFQ market makers, internalization, cross-chain routes). The protocol enforces user constraints (such as minimum output), but it does not enforce a single routing approach.
1inch’s Fusion mode is an intent-based swap where a third party resolver fills the order and covers gas, typically using a Dutch auction curve that decreases the rate until a resolver can fill profitably.
The common theme is that solvers/resolvers earn a spread by delivering execution while taking on operational risk.
RFQ (Request for Quote) is a pattern where the system asks market makers for firm quotes off-chain, then settles the chosen quote on-chain.
0x’s order documentation distinguishes order types including RFQ orders, which are designed for real-time quotes from market makers.
RFQ is attractive because it can:
RFQ also creates a dependency on the quoting set. If quote access is limited to a small set of market makers, execution quality becomes a function of that set’s competitiveness and uptime.
Intent-based systems are often described as “MEV protection.” That is directionally true for common user harms (sandwiching), but MEV does not disappear. It moves.
When orders are not broadcast as public swap transactions with loose slippage, classic sandwich attacks become harder. Many intent systems also enforce a strict minimum output, making slippage extraction bounded.
Order flow auctions create a different MEV arena:
From a user perspective, the relevant question is whether competition is broad enough that the solver margin is minimized.
Some intent systems are permissionless at the protocol level but have permissioned or curated components (whitelisted quoters, solver allowlists, or curated access to order flow). Even when the on-chain contracts are open, the off-chain coordination layer can concentrate power.
This concentration can show up as:
In direct swaps, the user controls inclusion by paying gas. In intent-based trading, the user depends on a solver to pick up the order. If the solver market is thin, orders can time out, especially during volatility.
Cancellation features vary by protocol. If cancellation requires an on-chain transaction, “gasless” systems still require users to hold gas when they need to stop an order.
| Model | Who Routes | How Competition Works | Typical Benefits | Typical Risks |
|---|---|---|---|---|
| Batch auction (CoW-style) | Solver | Solvers submit batch solutions; best surplus wins | Netting, better execution, MEV reduction | Solver concentration, batch timing effects |
| Dutch auction (UniswapX/1inch-style) | Filler/Resolver | Rate decays until filled; fillers compete on timing and routing | Gasless UX, price improvement, reduced sandwiching | Exclusivity windows, liveness dependency |
| RFQ-heavy execution | Market maker | Off-chain quotes, on-chain settlement | Deep liquidity, strong pricing on size | Quoter set concentration, quote availability |
These are not mutually exclusive. Many systems mix RFQ with AMM routing and additional constraints.
Intent systems are safest when the user’s constraints are tight and easy to enforce.
The minimum received amount is the user’s hard loss bound for price movement and solver margin. A long deadline increases the time window in which market conditions can drift.
A permissionless filler market tends to tighten margins when there are many capable participants. A permissioned solver set can still deliver quality execution, but it shifts trust from “open competition” to “operator selection.”
Fees can be charged as explicit protocol fees, solver spreads, integrator fees, or implicit price improvement capture. The cleanest user experience is when the only enforced rule is minimum output and the fill is free to compete.
Useful questions:
Intent systems that settle atomically on-chain with strict constraints are easier to reason about than systems with complex multi-step custody. A non-custodial claim is strongest when user funds are only transferred at settlement and only under the signed constraints.
Intent-based trading replaces direct routing with outcome constraints and solver competition. That shift can reduce common MEV harms and deliver better execution, especially when solvers are numerous and competing aggressively.
The tradeoffs are real: MEV moves into auction dynamics, liveness depends on solver markets, and some designs concentrate off-chain control even when on-chain contracts are permissionless. Users reduce exposure by keeping minimum outputs and deadlines tight, understanding who can fill, and preferring systems where competition is broad and settlement is strictly constrained by the signed intent.
The post Intent-Based Trading Explained: Solvers, RFQ Fills, and the MEV Tradeoffs appeared first on Crypto Adventure.
