Most people who trade on decentralized exchanges do not realize they are acting as their own market makers. Intent-based DEXs flip this model. Instead of interacting with a liquidity pool directly, you sign a message describing what you want, and a network of solvers competes to give it to you at the best price. The shift from execution to intention is the most significant change in DEX architecture since the automated market maker was introduced.
Summary
- Automated market makers changed decentralized trading by replacing order books with liquidity pools, but the model has structural costs. Slippage on large orders, impermanent loss for liquidity providers, and vulnerability to MEV extraction are all consequences of a design that requires traders to execute against a fixed pricing curve rather than competing for the best available price across all sources.
- Intent-based DEXs separate what a trader wants from how it gets executed. A trader signs an off chain message (an intent) specifying the token they want to sell, the token they want to receive, and the minimum acceptable price. The intent does not specify a route, a liquidity source, or a gas payment. All execution details are delegated to third party solvers.
- Solvers are the competitive layer in intent-based systems. They receive batches of user intents and compete to fill them at the best price. Solvers can route through multiple AMM pools, use their own private inventory, access centralized exchange liquidity, or match orders directly against each other. The solver that offers the best execution wins the right to fill the order.
- UniswapX uses Dutch auctions to discover prices, starting above market price and declining over time until a solver fills the order. CoW Protocol batches multiple intents together and finds coincidences of wants, matching opposing trades directly before routing remainders to on chain liquidity. 1inch Fusion delegates execution to resolvers who compete on price within a time window.
- MEV protection is a core benefit. In the AMM model, a pending transaction sits in the public mempool where searchers can sandwich it. In the intent model, the signed message is off chain and invisible to mempool searchers. Solvers execute the trade privately, and the Dutch auction mechanism ensures that any surplus value flows back to the trader rather than being captured by block builders.
The standard explanation of intent-based DEXs starts by contrasting them with AMMs, which is the right framework but the wrong emphasis. The interesting question is not what intents are. It is why AMMs, which solved the liquidity problem that made early DEXs unusable, are now being partially replaced by a model that reintroduces intermediaries.
The answer is that AMMs solved the wrong problem too well. They made it trivially easy to provide liquidity to any token pair, which is why decentralized trading volume grew from nearly zero to trillions of dollars. But they made it structurally expensive to trade, because the pricing curve that makes AMMs work also makes them predictable, exploitable, and wasteful for any trade larger than the pool can absorb without significant price impact.
Intent-based DEXs are not a rejection of AMMs. They are a layer on top of them that lets traders access AMM liquidity without bearing the full cost of AMM mechanics. Understanding how requires looking at what AMMs actually do and what they force traders to accept.
How AMMs price trades and why it costs you
An AMM uses a mathematical formula to determine the price of a trade. The most common is the constant product formula, x * y = k, where x and y are the reserves of two tokens in a pool, and k is a constant. When you buy token A from the pool, you add token B, and the formula adjusts the price so that the product of reserves stays the same.
This is elegant but expensive. The price you receive depends entirely on the size of your trade relative to the pool. A small swap barely moves the price. A large swap pushes the price along the curve, and the further it moves, the worse the effective price becomes. This is slippage, and it is not a bug. It is the mechanism by which AMMs prevent the pool from being drained.
The second cost is less visible. Every trade you submit to an AMM enters the public mempool before it is included in a block. Specialized searchers monitor the mempool for profitable trades and insert their own transactions before and after yours, a practice known as sandwich attacking. The searcher buys the token just before your transaction, pushing the price up, lets your trade execute at the worse price, then sells immediately after for a profit. You receive a worse price, the searcher captures the difference, and the AMM pool is none the wiser.
The third cost falls on liquidity providers. When the true market price of a token moves but the AMM pool has not yet been arbitraged to reflect it, anyone trading against the pool gets a better price than the market, and the liquidity provider absorbs the loss. This is called impermanent loss, though the word impermanent is misleading because the loss becomes permanent any time the price does not fully revert.
These three costs, slippage, MEV extraction, and impermanent loss, are not incidental. They are structural features of the AMM design. Intent-based DEXs address all three.
What an intent actually is
An intent is a signed off chain message that says: I want to sell X amount of token A for at least Y amount of token B, and I am willing to wait until a specified deadline for it to be filled. That is it.
The intent does not specify which liquidity pool to use, what route to take, whether to split the trade across multiple venues, or how to pay for gas. It does not interact with any smart contract at the time of signing. The trader tokens remain in their wallet, authorized for spending by the intent protocol settlement contract but not moved until a solver fills the order.
This separation is the core innovation. In the AMM model, the trader decides how to execute and bears the consequences of that decision. In the intent model, the trader specifies the outcome they want and delegates execution to a competitive marketplace of solvers who are better equipped to find optimal routes.
The distinction matters because it shifts the information asymmetry. In the AMM model, the trader broadcasts their intention to the entire network and anyone can exploit it. In the intent model, the trader signed message goes to a restricted set of solvers who compete on execution quality rather than on the ability to extract value from the trader transaction.
How solvers compete for your order
Solvers are the execution engines of intent-based DEXs. They are typically professional market makers, proprietary trading firms, or algorithmic strategies that specialize in finding the best possible execution for a given intent.
When a solver receives a batch of intents, it has several options for filling them.
Direct matching. If one user wants to sell ETH for USDC and another wants to sell USDC for ETH, the solver can match them directly without touching any liquidity pool. This is a coincidence of wants, and it produces the best outcome for both traders because there is no slippage, no pool fees, and no MEV exposure. CoW Protocol pioneered this approach and optimizes for it by batching intents before routing.
AMM routing. The solver can route the intent through one or more AMM pools, exactly as a traditional DEX aggregator would. The difference is that the solver pays the gas and absorbs the routing complexity. If splitting a trade across Uniswap v3 on Ethereum, SushiSwap on Arbitrum, and a Curve stableswap pool produces a better price than any single venue, the solver executes that multi-hop, multi-chain route.
Private inventory. Some solvers hold their own token balances and can fill orders immediately from inventory, bypassing on chain liquidity entirely. This eliminates slippage and MEV risk but requires the solver to manage their own market risk.
Centralized exchange liquidity. Sophisticated solvers maintain positions on centralized exchanges and can offer on chain users prices that reflect centralized exchange depth. This is controversial because it reintroduces centralized infrastructure, but it objectively produces better prices for large trades.
The competition between solvers is what makes the model work. A single solver with a monopoly would extract value the same way MEV searchers do. Multiple solvers competing on price push execution quality toward the theoretical optimum. The protocol role is to ensure this competition is fair and that solvers cannot collude or manipulate the auction.
UniswapX, CoW Protocol, and 1inch Fusion compared
The three leading intent-based protocols take different approaches to the same problem.
UniswapX uses a Dutch auction mechanism. When a user submits an intent, it starts at a price slightly better than the current market rate and declines over a defined time window (typically 12 to 60 seconds). The first solver willing to fill at the current auction price wins the order. This mechanism ensures that solvers compete on speed and efficiency. If the market price of the trade is $1,000 and the auction starts at $1,005, a solver that can execute at $1,003 will fill the order early and capture a $3 surplus. If no solver fills it quickly, the price declines until it reaches a level where execution is profitable. The user is guaranteed at least their minimum specified price.
CoW Protocol (Coincidence of Wants) takes a batch auction approach. Instead of processing intents individually, CoW collects all intents within a batch window (approximately 30 seconds) and runs a solver competition for the entire batch. Solvers propose settlement solutions that maximize the surplus for all users in the batch. The protocol selects the solution that produces the best aggregate outcome. This batching allows CoW to find coincidences of wants across multiple intents, eliminating the need for on chain liquidity entirely when opposing orders exist.
1inch Fusion uses a resolver network similar to UniswapX filler network but with different auction dynamics. Users submit intents that are filled by professional resolvers within a time window, with the price declining from a starting rate. 1inch advantage is its existing aggregation infrastructure, which gives resolvers access to the broadest set of on chain liquidity sources across multiple chains.
Each model has tradeoffs. Dutch auctions provide fast execution but depend on an active solver network. Batch auctions maximize surplus extraction but introduce latency. Resolver networks leverage existing infrastructure but concentrate execution among a small number of professional firms.
The volume flowing through intent-based protocols has grown rapidly since 2023. UniswapX processed over $30 billion in cumulative volume within its first year of operation. CoW Protocol regularly handles over $1 billion in weekly volume across Ethereum and Gnosis Chain. 1inch Fusion has become the default execution mode for a significant share of 1inch users, replacing the traditional aggregator routing path.
These numbers are significant but still represent a fraction of total on chain trading volume. The majority of DEX trades still execute directly against AMM pools. The transition from AMM-first to intent-first trading is gradual, driven by the size of savings on each individual trade rather than by a sudden platform migration. Users who trade frequently in sizes above $1,000 tend to discover the benefits of intent-based execution organically. Users who trade small amounts infrequently may never notice the difference.
The gas question: who pays and how
One of the most immediately noticeable differences for users is that intent-based swaps are often gasless. The trader signs an off chain message, which costs nothing. The solver submits the on chain transaction and pays the gas fee, pricing it into the execution.
This is not free gas. The cost is embedded in a slightly worse price. If the gas fee for a swap is $2, the solver will offer a price that is approximately $2 worse than the theoretical best execution. For a $100 swap, this 2% embedded gas cost might be worse than paying gas directly. For a $10,000 swap, the 0.02% embedded cost is negligible.
This dynamic creates a natural segmentation. Intent-based DEXs tend to offer better value for larger trades where the gas cost per dollar traded is small and the slippage savings from solver competition are large. Traditional AMMs remain competitive for small trades where gas optimization matters more than execution quality.
The gas abstraction also opens intent-based DEXs to users who do not hold the native token of the chain. A user holding only USDC on Ethereum can trade through an intent protocol without first acquiring ETH for gas, removing a significant onboarding friction that AMMs impose.
The broader implication of gasless execution extends to onboarding new users into decentralized finance. Traditional AMMs require users to hold the native gas token of whatever chain they are trading on, which creates a circular problem: you need ETH to trade on Ethereum, but you might be arriving with only USDC from a centralized exchange. Intent-based protocols dissolve this friction entirely. The solver handles the gas, the user handles the decision about what to trade, and the infrastructure disappears into the background where it belongs.
What this does not cover
This article does not cover the game theory of solver competition in depth. The mechanism design that prevents solver collusion, ensures fair ordering within batch auctions, and handles adversarial behavior is a complex topic with active research and evolving implementations.
This article does not cover cross chain intent protocols. UniswapX and others are building toward cross chain swaps where an intent on Ethereum can be filled with liquidity on Arbitrum or Base, but these systems are still early and introduce bridge security considerations that deserve separate treatment.
This article does not address the regulatory implications of solver networks. Whether solvers constitute broker-dealers, whether intent protocols are exchanges under securities law, and how these classifications affect protocol design are open questions that will be shaped by jurisdictional decisions over the next several years.
Practical checks before using an intent-based DEX
Check the solver count. An intent protocol with two active solvers offers less competitive execution than one with twenty. Most protocols publish solver activity data. More solvers means more competition, which means better prices.
Compare actual execution against a baseline. Intent protocols publish settlement data on chain. Compare the price you received against the spot price at the time of your intent submission. This tells you whether the solver competition is producing genuine savings or whether a thin solver market is extracting value.
Understand the expiry window. Your intent has a deadline. If no solver fills it before the deadline, the intent expires unfilled and your tokens stay in your wallet. A shorter deadline reduces the risk of price movement against you but limits the time solvers have to find the best execution.
Verify the settlement contract. Intent protocols use a smart contract to settle trades atomically. This contract has been approved to spend your tokens. Verify that you are interacting with the official, audited settlement contract and not a phishing contract that mimics the approval flow. Wallet security fundamentals apply here as they do everywhere.
Know when AMMs are still better. For very small trades, the gasless benefit of intent protocols may not offset the solver embedded gas cost. For illiquid token pairs where few solvers are active, a direct AMM swap may execute faster and at a comparable price. Intent protocols are not universally superior. They are superior in specific conditions that depend on trade size, token liquidity, and solver competition.
What is an intent-based DEX?
An intent-based DEX is a decentralized exchange where traders sign off chain messages describing what they want to trade rather than executing transactions directly against a liquidity pool. Professional solvers compete to fill these intents at the best possible price, routing through AMMs, private inventory, or direct order matching. The trader specifies the outcome; the solver handles the execution.
How do intent-based DEXs differ from AMMs?
AMMs require traders to execute directly against a liquidity pool using a mathematical pricing curve, which causes slippage on large orders and exposes trades to MEV extraction. Intent-based DEXs separate the trader desired outcome from the execution method. Solvers compete to fill intents using any available liquidity source, often producing better prices because they can access multiple venues and match orders directly without pool interaction.
What is a solver in intent-based trading?
A solver is a third party entity that competes to fill user intents at the best price. Solvers can be market makers, algorithmic trading firms, or automated strategies. They receive batches of user intents and determine the optimal execution path, which might involve routing through multiple DEX pools, using their own inventory, or matching opposing orders directly. The solver that offers the best price wins the right to fill the order.
Are intent-based swaps really gasless?
The trader does not pay gas directly because they sign an off chain message rather than submitting an on chain transaction. However, the solver who fills the intent pays gas to settle the trade on chain and embeds this cost in the execution price. The effective cost is typically lower than paying gas directly because solvers can batch multiple orders and optimize gas usage, but it is not zero.
What is a coincidence of wants?
A coincidence of wants occurs when two or more traders want to make opposing trades. If one user wants to sell ETH for USDC and another wants to sell USDC for ETH, a solver can match them directly without using any liquidity pool. This produces better prices for both traders because there is no slippage, no pool fees, and no MEV exposure. CoW Protocol specifically optimizes for finding these matches within batches of user intents.
How do intent-based DEXs protect against MEV?
In the AMM model, pending transactions sit in the public mempool where MEV searchers can sandwich them. Intent-based DEXs keep the signed intent off chain, invisible to mempool searchers. Solvers execute trades through private channels, and auction mechanisms like Dutch auctions ensure that any surplus value from execution flows back to the trader rather than being captured by block builders or arbitrageurs.
What happens if no solver fills my intent?
If no solver fills your intent before the specified deadline, the intent expires and your tokens remain in your wallet untouched. You are not charged any fee for an unfilled intent because no on chain transaction was submitted. You can resubmit a new intent with adjusted parameters if the original one expires.
Which intent-based DEX should I use?
The best choice depends on trade size and priorities. UniswapX is integrated into the Uniswap interface and uses Dutch auctions for fast execution. CoW Protocol optimizes for coincidence of wants and tends to produce better results for common trading pairs during high volume periods. 1inch Fusion leverages the broadest aggregation across liquidity sources. For large trades where execution quality matters most, comparing settlement prices across all three on a specific pair gives the most useful answer. This is educational analysis, not investment advice.
Disclaimer: This article is for informational and educational purposes only. It does not constitute financial, investment, or legal advice. Cryptocurrency markets are volatile and carry significant risk. Always conduct your own research before making investment decisions. Published August 11, 2026.
Be the first to comment