If you are moving USDC, USDT, or USDG across chains, the Solver network sitting behind your intent protocol is the single biggest determinant of whether your transfer lands in six seconds or sits in limbo for six minutes. Solver networks stablecoin flows depend on are off-chain competitive markets: participants race to fill user intents, front their own capital on the destination chain, and claim the escrowed funds after proof. The more Solvers, the tighter the price. The more stablecoin-native the network, the less you pay in slippage and routing overhead. This guide compares the 8 Solver networks that matter most for cross-chain stablecoin transfers in 2026 on what their own documentation actually publishes: who fills the intent, how settlement is proven, and whether the design is stablecoin-specific. You will also see where each network sits on three decision dimensions, Solver depth, capital efficiency per Solver, and stablecoin specialization, so you can pick the right fit for payments, treasury, or trading.
​

A Solver network is the off-chain market that fills intents. When a user signs an intent, "I will pay 10,000 USDC on Base, I want 9,998 USDC on Arbitrum in under 15 seconds", the intent is broadcast to a mempool of Solvers. Each Solver decides whether to bid. The winning Solver fronts the destination funds instantly, then claims the source escrow once a settlement proof is posted. The user sees one atomic action. Under the hood, Solvers compete on price, speed, and reliability. This is the intents and solvers architecture that now underpins most modern cross-chain stablecoin UX. Some networks call their participants Fillers (UniswapX) or Relayers (Across); the mechanics are similar, but the terminology fractures the market. This guide uses "Solvers" throughout, following Eco's preferred terminology and the ERC-7683 cross-chain intents standard.
​

Every Solver network can be scored on three axes. First, active Solver count: more Solvers means more competition, tighter spreads, and higher liveness when one goes offline. Second, capital efficiency per Solver: a Solver running on a fast settlement layer with efficient rebalancing can recycle the same dollar across dozens of fills per hour; a Solver stuck behind slow bridges can only fill one at a time. Third, stablecoin-native vs generic: a network designed for stablecoin flows uses 1:1 matching, oUSDT/USDC-equivalent routing, and avoids unnecessary swap hops, while a generic network treats USDC like any ERC-20 and quotes you DEX slippage. If you want more background on how these competitive markets clear trades, the Paradigm essay on intent-based architectures is still the canonical read.
​

Every column above restates what each protocol's own documentation says, checked on September 15, 2026. Two of these networks publish a fill-time figure: Across (docs.across.to) and Wormhole Settlement's Mayan Swift route (wormhole.com docs). None of the eight publishes a live count of active Solvers or a stablecoin share of volume, so neither figure appears here.
​

Eco's Solver network is the only entry on this list that is explicitly stablecoin-native. Solvers on Eco Routes compete to fill intents denominated in USDC, USDT, USDC.e, oUSDT, USDT0, USDbC, and USDG. Eco's developer docs describe Routes as covering more than 16 chains and more than 240 directional pairs, with typical fulfillment in 20 to 40 seconds (docs.eco.com). Because Eco's intent schema is 1:1 stablecoin matched rather than generic ERC-20 swap, Solvers quote without DEX slippage layered on top. Developers publish intents via the Routes CLI or API, see how to publish a cross-chain intent. For deeper architecture, Eco's writeup on blockchain solver netting covers how capital gets recycled across fills.
​

Across calls its Solvers "Relayers." The network pioneered the optimistic-fill model where a Relayer fronts destination funds immediately and proves the source deposit asynchronously through UMA's optimistic oracle. Across documents fills of about 2 seconds and says running a Relayer is permissionless, so anyone can operate one (docs.across.to). That makes it the fastest published figure on this list. The design is not stablecoin-native, it supports ETH, WBTC, and other assets. Across does not publish a Relayer count or a stablecoin share of volume. Its Relayer fee documentation explains what Relayers are compensated for. For a broader view of how Across compares with other cross-chain intent protocols, start there.
​

UniswapX uses "Fillers", functionally the same role as Solvers. The network launched with a Dutch-auction mechanism where quoted prices decay over a short window, giving Fillers a deterministic way to win competitive intents. UniswapX is same-chain first; the cross-chain version rolled out later and is still scaling stablecoin volume. The network is generic rather than stablecoin-specific: ETH, wstETH, and long-tail tokens trade alongside dollars. Uniswap publishes neither a Filler count nor a median fill time, and all orders are backstopped by the Uniswap Smart Order Router, which forces Fillers to compete with the Uniswap pools themselves. If you want more context on the protocol design, Eco's breakdown of UniswapX and gasless swaps walks through the Dutch-auction mechanics. The Uniswap Foundation's UniswapX protocol launch post is also worth reading.
​

CoW Protocol was the first to productize the term "Solver." CoW Solvers bid in discrete batch auctions, and the Solver whose solution generates the greatest surplus for the batch wins the right to settle it, crossing user orders against each other before hitting external liquidity. The batch model trades speed for price improvement and MEV protection, which matters for large stablecoin swaps where a few basis points of savings outweigh a short wait. CoW publishes no Solver count, no batch latency figure, and no stablecoin share of volume. Eco's reference piece on CoW Swap and intent-based DEX trading explains the batch auction in detail. The CoW Protocol Solver docs describe Solvers as independent, bonded participants competing in that auction, and note that anyone able to code an optimization algorithm can run one.
​

1inch calls its Solvers "Resolvers." Fusion is a Dutch-auction intent layer bolted onto the existing 1inch aggregation router, so Resolvers can route fills through 1inch's aggregated DEX liquidity on the destination chain. The Resolver pays the onchain gas, so the user never needs a native balance, and the order rate decays through a Dutch auction until a fill becomes profitable. 1inch publishes no Resolver count and no median fill time. Fusion+ extended the model to cross-chain flows; documentation is in the 1inch Intent Swap docs. Like UniswapX, Fusion is generic-first, stablecoin routing works well but is not the design center of gravity. For traders weighing Fusion against cross-chain liquidity protocols, Eco's 2026 comparison covers the fee curves.
​

Aori is a newer cross-chain intent-settlement layer. Its protocol docs describe a four-state order lifecycle: tokens lock on the source chain, a Solver fills on the destination, and a LayerZero message confirms the fill and unlocks the input tokens for the Solver. It supports ERC-20 tokens and native ETH rather than a stablecoin-only schema, and publishes no Solver count, fill time, or stablecoin share of volume. Eco's writeup on Aori intent settlement covers the architecture. If you are an institutional desk evaluating institutional stablecoin RFQ options, Aori deserves a look alongside Eco Routes. Capital efficiency per Solver is strong because each Solver knows the counterparty set.
​

deBridge DLN (Decentralized Limit orders Network) runs a Solver network that fills cross-chain limit orders. Its docs describe DLN as 0-TVL infrastructure: there are no liquidity pools, trades settle asynchronously through a self-organized liquidity network, and orders fill with zero slippage at any size. The deBridge model leans on its own messaging layer for settlement proofs rather than an external oracle, which reduces latency at the cost of a smaller trust-minimization surface than optimistic systems. deBridge publishes no Solver count, fill time, or stablecoin share of volume. Documentation is in the deBridge DLN technical docs. For stablecoin-first flows, capital efficiency is middle-of-the-road because Solvers must hold inventory across a long tail of asset pairs, not just dollar tokens. This makes deBridge a solid choice for generic cross-chain but a suboptimal one for pure stablecoin rail use.
​

Wormhole Settlement sits on top of the Wormhole message-passing layer and runs two routes. Mayan Swift uses fast off-chain auctions among what the docs call a curated set of solvers, with execution typically around 12 seconds; Mayan MCTP wraps Circle's CCTP for native USDC bridging and is slower because it waits on chain finality (wormhole.com docs). The generic-first design and the dependency on Wormhole Guardian signing make it a good fit for exotic cross-VM destinations where no other Solver network operates, but a weaker fit when your flows are USDC-on-EVM or USDT0 across L2s. Wormhole publishes no solver count or stablecoin share of volume. For most EVM and Solana stablecoin flows, earlier entries on this list will clear faster and tighter.
​

If your flows are stablecoin-only and you care about fill time and price, start with Eco and Across, the two networks on this list that publish a fill-time figure at all. If you are willing to trade speed for price improvement on large tickets, CoW's batch model earns the wait. If you want institutional RFQ with curated counterparties, Aori fits. If your destination chains include Sui, Aptos, or other non-EVM tails, Wormhole Solver may be your only option despite the weaker stablecoin share. For developers building against these networks, the choice often comes down to integration surface: Eco Routes exposes a declarative intent API where you state the outcome and the network picks the fill, while generic networks like deBridge require you to construct the hop graph yourself. For more on that distinction, see Eco's overview of what a blockchain intent solver does and the broader comparison of intent-based routing protocols.
​

Most teams do not integrate Solvers directly, they integrate with the intent protocol that sits in front of the Solver network. On Eco, the integration path is the Routes CLI and Routes API. A typical onchain flow looks like: sign an intent specifying source chain, destination chain, source token, destination token, and minimum receive amount; the Routes infrastructure broadcasts the intent; a Solver fills it on the destination chain; settlement is atomic or the escrow refunds. This replaces the traditional "bridge then swap then hope" pattern with a single declarative call. Eco publishes step-by-step docs at publishing an Eco Routes intent. The Routes CLI source on GitHub is the quickest way to test-fire an intent on testnet before production. If you are evaluating this model against hop-explicit routers, Eco's broader cross-chain liquidity protocols roundup is a useful companion read.
​

A Solver network is an off-chain competitive market of independent participants who compete to fill user intents. Each Solver fronts destination-chain funds immediately and claims the user's escrowed source funds after a settlement proof is posted. More Solvers means tighter prices and higher reliability. See the intents and solvers guide for the full mechanics.
​

They are the same role under different brand names. Eco and CoW Protocol use "Solvers." UniswapX uses "Fillers." Across uses "Relayers." 1inch uses "Resolvers." Each fills user intents by fronting destination funds. The terminology differs but the mechanic is the same competitive off-chain fill market, standardized by ERC-7683.
​

Eco Solver Network is the only one of the eight with a schema designed for 1:1 dollar matching rather than generic ERC-20 swap, and Eco's docs put typical Routes fulfillment at 20 to 40 seconds across more than 240 directional pairs (docs.eco.com). Across is the closest competitor and documents fills of about 2 seconds (docs.across.to), though it is a general-asset network rather than a stablecoin-specific one. None of these networks publishes an active Solver count or a stablecoin share of volume, so no honest ranking on those two axes is possible.
​

Solvers capture the spread between the fill price quoted to the user and the cost of sourcing destination liquidity and proving settlement. They also earn gas rebates or protocol incentives on some networks. In a deep Solver market the spread compresses, which is why active Solver count matters, it pushes prices toward the user.
​

Generally yes for stablecoin flows. A bridge locks source funds and mints a wrapped representation on the destination, creating a custody surface that can be exploited. A Solver network escrows the source and releases it to the Solver after a cryptographic proof of destination fill, users either receive the exact expected funds or the intent expires and source funds refund. See Eco's comparison of cross-chain intent protocol options for more.
​