For most of DeFi's history, moving a token from one place to another meant walking a routing table. A smart order router priced every hop, stitched together pools, and hoped nothing moved while the transaction was in flight. Solver networks crypto users are starting to prefer flip that model: users sign what they want, and a competitive marketplace of solvers bids to deliver it. The winner is whoever can settle fastest at the best price — the routing table is nobody's problem but theirs.
​

This article explains what a solver is, why solver competition produces better execution than static routers, how settlement and MEV dynamics work under the hood, and how live solver networks like Eco Routes turn the model into a production-grade execution layer. By the end you should be able to read a new intent protocol's architecture and tell whether the solver economics actually hold up.
​

A solver is an independent actor — usually an automated market-making firm or a searcher — that competes to fulfill user-signed intents. Instead of the user specifying the path ("swap on Uniswap, bridge through X, unwrap on Y"), the user specifies the outcome: "I have 10,000 USDC on Arbitrum; I want at least 9,990 USDT on Base, settled within 30 seconds." A solver reads that intent from a public order flow, prices it against its own inventory and liquidity sources, and — if the math works — commits to filling it. The formal literature on this model goes back to the Anoma intents thesis, which frames intents as partial commitments that only become transactions when a solver matches them.
​

The intent layer takes care of the cryptographic guarantees. Eco's pillar explainer on intents and solvers walks through the full handshake: the user signs an intent, the solver posts a collateralized fill, and atomic settlement either completes or reverts with no in-flight funds stuck. The solver eats the risk of timing, slippage, and gas; the user just sees a confirmation.
​

What makes a blockchain intent solver different from a market maker is the breadth of the problem space. A solver on a modern network prices simultaneously across onchain AMMs, private RFQ desks, CEX inventory, and cross-chain rails. Their competitive edge is not a single pool — it's how well they route inside the network.
​

The traditional smart order router is deterministic. It takes a quote, walks a precomputed graph of pools, and fires a transaction. It is a good solution when liquidity lives in a handful of AMMs on a single chain. It breaks down the moment you care about multiple chains, private venues, or latency-sensitive pricing — because the router cannot see into every venue, and it cannot negotiate.
​

A solver network replaces the router with an auction. Paradigm's original intents thesis captures the shift concisely: express the outcome, let a market bid for it. What this produces in practice is not just better prices — it's better worst-case prices, because the losing solvers still have to have been willing to fill the order. In a static router you get the top-of-book mid; in a solver auction you get the second-best solver's bid as a floor.
​

Eco's guide to intent-based routing protocols goes deeper into how competitive bidding compresses spreads on cross-chain flows in particular. On a single chain, the gap between a router and a solver is a few basis points. On a cross-chain stablecoin swap, the gap is often 20-40 basis points — because routers can't model bridge liquidity, but solvers can prefund and rebalance against it. The a16z crypto research on intents makes a similar point from a slightly different angle: static routers are optimizing a graph they can see, while solvers are optimizing over a much larger search space that includes private liquidity.
​

A clean solver network has four stages — intent origination, auction, commitment, and settlement — and each stage has its own integrity guarantees.

The elegant part is that the user's perspective collapses all four stages into a single signed object. The complexity lives inside the solver. For a fuller picture of how settlement proofs and solver netting interact across thousands of concurrent intents, see Eco's writeup on blockchain solver netting.
​

Maximal extractable value is the elephant in the room for any intent protocol. The naive worry is that a solver, sitting between the user's signed intent and the onchain execution, is perfectly positioned to sandwich or back-run. The less obvious reality is that a well-designed solver network is actually better at MEV protection than a public mempool.
​

Three mechanisms make this work. First, the auction itself is an MEV-internalization engine — if a solver can extract X basis points of MEV, competing solvers will bid X back into the price to win the flow. Second, solvers run on private order flow channels that don't leak to the public mempool, similar to the protections Flashbots' protected RPC offers for direct swaps. Third, explicit MEV-sharing agreements — pioneered by CoW Swap's batch auction model and expanded in UniswapX's filler rebate logic — route captured MEV back to users as improved quotes. The SUAVE research from Flashbots sketches what it looks like to build a dedicated chain for this kind of auction.
​

This doesn't mean MEV disappears. It means MEV gets priced into the quote the user sees, instead of being silently extracted after the fact. The result is usually tighter effective spreads for the user and a healthier margin for the solver who wins on efficiency, not extraction.
​

Who funds a solver's bid? The answer is what separates sustainable solver networks from the short-lived ones.
​

A solver's economics have three inputs. Gross margin comes from the spread between the user's intent price and the best executable price the solver can find — across AMMs, RFQ, CEX inventory, and internal book. Gas and settlement costs are paid by the solver, not the user, which is why gasless intent flows feel magical: the solver is simply pricing gas into the spread. Risk capital is the collateral a solver has to lock during the commitment phase; its opportunity cost is a real line item.
​

Competitive pressure pushes gross margin down until only the most capital-efficient solvers survive. That sounds bad, but it's the point — the protocol wants survivors who can fill orders at the tightest viable spreads. Networks typically stabilize with 3-8 active solvers per lane. Fewer and the auction isn't competitive; more and the economics thin out to the point of dropout.
​

The second-order effect is specialization. Solvers tend to specialize by chain family (EVM vs SVM), by asset type (stablecoins vs majors vs long-tail), and by flow size (retail vs institutional RFQ). Eco's coverage of stablecoin liquidity networking explains how a solver-network backbone actually benefits from that specialization: the protocol aggregates diverse solvers into a single interface, and users never have to know who filled their order. The architectural parallel is equities execution — venue competition has long been understood to improve execution quality for end users, and the solver auction is the crypto-native analogue of that dynamic. The Jump Crypto research hub covers related microstructure work.
​

Eco Routes is the production intent execution layer for stablecoin flows. The network spans 15 chains — Ethereum, Optimism, Base, Arbitrum, HyperEVM, Plasma, Polygon, Ronin, Unichain, Ink, Celo, Solana, Sonic, BSC, and Worldchain — and supports USDC, USDT, USDC.e, oUSDT, USDT0, USDbC, and USDG. A developer writes an intent through the Routes CLI or API; the network handles solver selection, liquidity aggregation, and atomic settlement.
​

What makes Eco's solver network distinct operationally is the route configuration surface. Hyperlane Routes and Native Routes let an integrator pin a specific settlement rail per lane — useful when the compliance team requires a particular message channel, or when a treasury wants deterministic cost modeling. Solvers still compete inside each lane; the rail just defines the finality guarantee.
​

For developers, the fastest way to see the network in action is the walkthrough on publishing your first cross-chain intent. Sign once, watch multiple solvers bid in real time, and get a filled transaction confirmation in under a minute. If you've been stitching bridges and swap aggregators together, the unification is the point — one integration, one signed intent, one settled outcome.
​

If you're comparing solver networks as a protocol integrator or treasury team, there are five questions that matter more than marketing pages:

Static routing survives where liquidity is concentrated, latency is forgiving, and a single chain is enough. Every other execution problem — cross-chain, multi-venue, latency-sensitive, large-size — is converging on solver networks. The economic reason is simple: competitive auctions are a better price-discovery mechanism than precomputed graphs when the graph is too big to compute.
​

For builders, the practical consequence is that "routing logic" is being commoditized. You don't need to maintain a cross-chain router; you need an orchestration layer that knows how to sign intents and read settlement events. Eco is one way to get there; there are others. What's non-negotiable is recognizing that routing tables are yesterday's primitive.
​

What are solver networks in crypto? Solver networks are marketplaces where independent actors (solvers) compete in real-time auctions to fulfill user-signed intents. The user specifies an outcome; the network selects the best bid and guarantees atomic settlement. They replace static smart order routers with a competitive execution layer across chains and venues.
​

How do solvers make money? Solvers earn the spread between the user's intent price and the best executable price they can find across liquidity sources, minus gas, collateral opportunity cost, and rebalancing expenses. Competitive pressure compresses margins toward the most capital-efficient solvers.
​

Are solver networks safer than bridges? In well-designed networks, yes. Solvers post collateral that's slashed if they fail to deliver, so the user is made whole automatically. Traditional lock-and-mint bridges rely on message-passing trust assumptions with no automatic user remediation if something stalls.
​

What's the difference between a solver and a market maker? A market maker quotes continuously on a single venue. A solver prices a specific user intent across many venues — AMMs, RFQ, CEX, cross-chain rails — and commits to execution. Solvers are closer to cross-venue search firms than to classical MMs.
​

Do solver networks handle MEV? Competitive auctions internalize MEV into quotes — if a solver can extract X basis points, competitors bid X back into the price. Private order flow channels and explicit MEV-rebate mechanisms further compress extraction. See CoW Swap's MEV protection approach for a concrete example.
​