A request for quote, or RFQ, is a trading workflow where a taker asks one or more market makers for a firm, signed price on a specific size, then chooses the best quote and settles it. RFQ powers a growing share of onchain stablecoin flow, including USDC, USDT, PYUSD, and Ethena USDe trades. Total stablecoin market capitalization sat at $315.3B on June 5, 2026, per DefiLlama, and a large fraction of institutional-size swaps in that pool now clear through RFQ venues rather than constant-product AMMs.
​

This article explains what RFQ means in stablecoin markets, how a quote actually executes, the components involved, why stablecoin liquidity has gravitated toward this model, and where it sits next to AMMs and intent-based swaps.
​

RFQ is a bilateral price discovery method. A taker submits the pair and size they want to trade, makers respond with firm two-way or one-way quotes, and the taker either hits a quote or lets it expire. Unlike a central limit order book, prices are not standing and public. They are quoted on demand for a specific request.
​

The model originated on TradFi FX and bond desks, where dealers historically quoted institutional clients by phone, then Bloomberg, then electronic RFQ platforms. The shape carried over to crypto because much of stablecoin flow looks more like FX than like equities. Pairs are concentrated, size is chunky, and takers care about certainty of execution more than about resting depth.
​

A central limit order book (CLOB) posts bids and offers continuously. Anyone can match against them, and price-time priority dictates fills. An automated market maker (AMM) replaces the book with a bonding curve, and price is a deterministic function of pool reserves. RFQ does neither. There is no resting book and no curve. Each price exists only inside a short-lived signed quote.
​

That structure has tradeoffs. RFQ gives takers price certainty before they commit gas, and it lets makers price off their full inventory and hedging stack rather than a fixed pool. Order books and AMMs offer transparent, anyone-can-trade price discovery, at the cost of slippage on size. For background on the orderbook alternative, see the 0x orderbook docs.
​

An RFQ trade runs as a short request-response loop with an atomic onchain settlement at the end. The taker broadcasts a quote request to one or more makers. Each maker returns a signed quote that includes price, size, expiry, and a nonce. The taker selects a quote, submits it to the settlement contract, and the contract verifies the signature and transfers tokens.
​

The standard sequence looks like this. First, the taker (or a router acting on their behalf) sends a request specifying input token, output token, amount, and chain. Second, makers price the request against inventory, hedges, and risk limits, then return a signed payload. Third, the taker compares quotes and picks one. Fourth, the taker submits the signed quote, plus their own signature or transaction, to the venue's settlement contract.
​

Quote time-to-live (TTL) is usually short. A short TTL protects the maker from being arbitraged against stale prices when the underlying market moves. If the taker waits past TTL, the quote is no longer honored and the taker must request again. Hashflow's RFQ documentation walks through the signature scheme and quote lifecycle in detail.
​

Settlement is atomic. The contract pulls the taker's input tokens, pulls or pre-funds the maker's output tokens, and reverts the entire transaction if any leg fails. Because the maker signed the exact price, there is no slippage between quote and fill. Some venues also support gasless flow where the taker only signs and a relayer pays gas, recouped from the trade.
​

An RFQ stack has six recurring pieces. Together they take a stablecoin swap from intent to onchain settlement without resting liquidity sitting in a pool. Each piece can be operated by a different party, which is part of why RFQ designs vary across venues even when the underlying mechanic is the same.

The split matters because it shapes who bears what risk. Makers carry inventory and hedging risk. Solvers carry execution and competition risk. Takers carry only the residual risk that no maker responds in time, in which case the request expires and they retry.
​

Stablecoin swaps reward price certainty and punish slippage, and that profile fits RFQ better than AMMs. A treasury moving $5M from USDC to USDT does not want to learn the effective rate from a curve. They want a firm number before they sign, and they want the settled amount to match. RFQ delivers that. AMMs, even concentrated-liquidity ones, do not.
​

Stablecoin supply is dominated by a few issuers. As of June 5, 2026, USDT supply was $187.2B, USDC was $75.6B, PYUSD was $2.9B, and Ethena USDe was $4.5B, per DefiLlama. That concentration means a handful of pairs (USDT/USDC, USDC/USDe, USDC/PYUSD, and their cross-chain variants) carry most volume. Makers can warehouse deep inventory in those names and quote tight prices on demand, which is exactly what RFQ rewards.
​

MEV is the other driver. AMM swaps post intent to the public mempool, where searchers can sandwich or front-run. RFQ quotes are signed for a specific taker and a specific nonce, which makes the trade economically uninteresting to sandwich. Makers also internalize toxicity rather than passing it to LPs.
​

Cross-chain settlement amplifies the advantage. A taker who wants USDC on Base for USDT on Ethereum does not want to discover the cross-chain price by sweeping pools. An RFQ venue can return one signed quote that covers the entire path, with the maker handling rebalancing across chains in the background.
​

RFQ, AMM, and intent-based swaps each optimize for different things. AMMs win on long-tail assets and small sizes where bilateral quoting overhead is wasteful. RFQ wins on stablecoin and major-pair size where slippage and MEV dominate cost. Intent-based designs blur the line by letting solvers choose between them, including using RFQ quotes as one input.
​

The table below summarizes the typical tradeoffs on a USDC-to-USDT trade of meaningful size.

Several venues sit between categories. 1inch Fusion uses an intent model with resolvers that can source from AMMs or quote like makers. CoW Swap batches user intents and lets solvers compete, often pulling RFQ quotes from professional makers into the batch. 0x ships both an orderbook and an RFQ stack, and many wallets route to whichever returns a better price for a given trade.
​

Eco Routes occupies a related but distinct seat. It is a neutral cross-chain orchestrator that lets developers integrate stablecoin movement across chains without picking a single rail. It does not hold inventory, does not act as a market maker, and does not take principal risk. When useful for a developer's flow, RFQ-style price discovery sits upstream of the Routes settlement layer.
​

RFQ is not free of tradeoffs. The same features that produce price certainty also concentrate trust in a small number of professional makers. Takers should understand the failure modes before relying on RFQ for production flow, especially treasury and payments workflows where settlement guarantees matter.
​

Maker centralization is the first concern. A venue with five active makers is one outage away from a thin book. Quote staleness is the second. If TTLs are too long, makers either widen prices to compensate or pull quotes mid-flight. Last-look, a TradFi import where a maker can reject a fill after the taker accepts, exists on some venues and not others. Takers should confirm whether a venue is firm-quote or last-look before integrating.
​

KYC posture varies. Some institutional RFQ venues require taker onboarding and accredited-only access. Permissionless RFQ venues do not, but they may have geofencing or maker-side compliance gates that affect which addresses can be quoted. Settlement risk is generally low when contracts are audited and atomic, but cross-chain RFQ adds a bridge or messaging layer that introduces its own assumptions.
​

Finally, RFQ quote quality depends on competition. A single-maker venue, or a venue where one maker wins most flow, will tend toward wider spreads over time. Takers benefit from polling multiple venues and from solver layers that force makers to compete on every request.
​

The arc points toward convergence with intent-based designs and toward native cross-chain RFQ. Intents already swallow RFQ as one of several pricing inputs. Solvers in CoW Swap, 1inch Fusion, and similar systems will source from AMMs, private pools, and RFQ quotes within the same batch, and the user sees only the cleared outcome.
​

Cross-chain is the second vector. As stablecoin movement increasingly spans Ethereum, Base, Solana, Arbitrum, and emerging chains, RFQ venues are quoting cross-chain pairs natively rather than chaining single-chain swaps. Settlement typically relies on cross-chain messaging and issuer-native rails to clear the underlying inventory.
​

The third vector is agentic commerce. As AI agents start to settle payments and treasury actions onchain, they need price certainty and a clear pre-trade quote. RFQ fits that shape better than blind AMM swaps. An agent can request a quote, evaluate it against policy, sign once, and settle, with no mid-trade surprises. That is the workflow stablecoin payments and treasury stacks are building toward.
​

Eco Routes contributes to that stack as a neutral cross-chain orchestration layer. Developers integrate once across markets, and the underlying pricing layer, whether RFQ, AMM, or intent solver, sits behind a stable interface. The institutional value prop is one integration across markets, with neutrality on which rail or pricing mechanic clears any given trade.
​

Methodology: stablecoin supply and market cap figures sourced from DefiLlama snapshot dated June 5, 2026. Venue mechanics sourced from each protocol's public documentation as cited inline.
​