Big crypto orders don’t just hit a pool and hope. They hunt for liquidity across AMMs, RFQ market makers, and sometimes private lanes most people never see. If you’ve ever tried to move size
Big crypto orders don’t just hit a pool and hope. They hunt for liquidity across AMMs, RFQ market makers, and sometimes private lanes most people never see. If you’ve ever tried to move size onchain and watched price wiggle away from you, this is for you.
We’ll break down how onchain RFQ actually works, how it stacks up against AMMs, where aggregators source fills, and the traps that clip big orders. I’ll also point to fresh datasets that trace the exact path your swap took, so you can verify where size really came from.
Why now? Onchain liquidity is maturing fast. Tokenized Treasuries and money market funds are now a huge footprint on public chains, a sign that bigger players are comfortable parking size onchain — and they expect pro-grade execution when they trade.
Large traders increasingly use onchain RFQ (request-for-quote) to get firm quotes from market makers and minimize price impact, while AMMs remain the default for smaller or long-tail tokens. Aggregators route between the two based on size, depth, and slippage settings, often splitting the order. In short: AMMs price continuously, RFQ quotes discretely; whales blend both.
RFQ is simple in spirit: you ask a network of market makers for a quote; they answer with prices; you pick the best and settle onchain. The wrinkle is coordination. Quotes might be computed offchain by PMMs (professional market makers) or solvers, then committed onchain via a small contract call. That’s why you’ll hear people say RFQ is hybrid. Discovery offchain, settlement onchain.
For a large trade, you or an aggregator package the order context: token pair, size, side, max slippage, and sometimes hints about hedging venues. Makers respond with a price that factors in their inventory, hedging routes (CEX, perps, correlated pairs), and the gas they’ll pay. If you accept, the maker executes the settlement transaction or signs a message you can settle yourself.
Good RFQ flows also include protections: quote expiry (seconds), match-ahead rules to prevent sandwiching, and sometimes private relay submission so the quote isn’t broadcast to searchers before it lands in a block.
AMMs price along a curve. Push size through and you slide down that curve. RFQ makers don’t reprice on a curve; they price your exact order, then try to hedge it. That difference matters when you’re trading size in thin markets.
Here’s a tight comparison that captures the main trade-offs:
Topic AMM Onchain RFQ Price formation Continuous curve from pool reserves Discrete quotes from makers/solvers Slippage on large orders Rises nonlinearly with size Embedded in quote; often lower for big tickets MEV exposure Higher if public mempool Often reduced via private lanes or committed quotes Settlement Direct pool swap Maker or solver contract executes onchain Counterparty Pool LPs, permissionless Named or whitelisted PMMs/solvers Long-tail tokens Often the only route May be unavailable or wide spreads Privacy Public unless using private RPC Negotiation offchain; onchain settlement visible Operational overhead Low, click-and-swap Quotes, expiries, signatures, and routing logic
In practice, aggregators blend both. They’ll test a small “warm-up” amount on AMMs to probe depth and price, ping RFQ makers for firm size, then split the order. You get a composite fill that hides a lot of moving parts.
Not every big trade needs to be chopped up. If the pool is deep and the pair is blue-chip, a single bite might be fine. But most of the time, splitting helps reduce impact, smooth execution through volatility, and keep your intent less obvious.
Here’s a quick checklist I use before slicing an order:
Time-weighted strategies (TWAP) still matter onchain. Many solvers can schedule chunks across blocks while monitoring quotes. Just set your max duration so you’re not chasing a moving target for hours.
Aggregators watch two streams: public AMM quotes and private RFQ responses from makers. They simulate fills with gas, route risk, and slippage. If an RFQ quote beats the best simulated AMM path after costs, they’ll take it. If the RFQ is stale or too wide, back to pools they go.
We can actually see this distinction now in public data. Allium’s DEX Analytics schema surfaces separate views for dex.trades (AMM pool fills) and dex.aggregator_trades (fills routed via aggregators, intent systems, and RFQ/PMM liquidity), which lets researchers tell how a swap was really executed (DEX Analytics (Allium)). Their Orderflow Explorer maps the full path of a swap across EVM chains — frontends, meta-aggregators, solvers or PMMs, then the final AMM pools if used — a clean way to trace where large trades actually sourced liquidity (DEX Analytics (Orderflow Explorer)).
It’s not just execution nerds looking. Allium was also cited as the primary on-chain data source in Bitwise’s inaugural Q3 2026 Staking Report, a small but telling sign that mainstream research is leaning on this type of orderflow clarity (Allium (Press & News listing / The Onchain Query)).
Pro tip: If an aggregator lets you, log or export the route. Later, compare it to Allium’s path view to verify whether your price came from AMM depth, a solver’s RFQ quote, or both. It’s the easiest way to sanity-check execution quality.
For AMMs, the main pain is MEV if you’re broadcasting a juicy swap. Sandwiching can widen your effective cost. Private RPCs and protected builders help, and many frontends now default to them for bigger trades. RFQ paths often push settlement via private lanes by design, which takes a lot of the sting out.
Gas is more straightforward. A single RFQ fill can be cheaper than multiple hops across pools, but not always. Hops and approvals add up. On volatile days, a maker might factor gas into their quote, so it’s reflected in the price you see. Always compare net-out, not just the headline spot.
Settlement risk is rare but real. With AMMs, you face smart contract risk and pool price movement between submission and inclusion. With RFQ, you add quote expiry and counterparty behavior. If a maker’s hedge slips, they can cancel or requote. Tight expiries and clear failure rules help, and good aggregators enforce them automatically.
Short answer: usually, yes, especially for size and for tokens tied to deeper macro markets. Tokenized real-world assets have become a serious footprint: as of mid-July 2026, Allium’s cross-chain RWA dataset shows $29.5 billion in tokenized RWAs, with $16.3 billion in Treasuries and money-market funds alone (Allium (The Onchain Query)). That liquidity doesn’t all sit in AMM pools; a lot is managed by professional desks who prefer firm quotes and private hedges.
If you’re rotating stablecoin treasuries, rebalancing RWA-backed vaults, or moving between L2s, RFQ makers can internalize some of that flow or hedge across CEX perps and SoFi-like rails faster than an AMM path can handle, particularly when rates or funding are jumpy.
That said, AMMs still dominate long-tail and experimental assets, and they’re great for immediate, permissionless access. Blue-chip majors increasingly see RFQ blends because the cost to move size quietly is just lower.
Cover image for Allium’s 'Mapping the Flow of DEX Swap Execution' — points to the Orderflow Explorer used to trace whether large swaps route to AMM pools or are filled by RFQ/PMM liquidity (useful as the entry visual to the interactive orderflow data). — Source: Allium (Orderflow Explorer blog post)
Two options. The quick one: check your aggregator’s route details. Many show which pools and which maker contract filled you. The thorough one: use an orderflow explorer that reconstructs the path.
Allium’s Orderflow Explorer does this across multiple EVM chains, stitching the frontend, any meta-aggregators, the solver or PMM that supplied a quote, and whether part of the order ultimately settled against an AMM pool (DEX Analytics (Orderflow Explorer)). Because their schema cleanly splits AMM pool events from aggregator/intent/RFQ events, it’s easier to attribute the price you got to the real liquidity source (DEX Analytics (Allium)).
Why care? If your best fills consistently come from a specific maker, you might optimize by expanding allowlists or pre-approvals with that counterparty. If they’re mostly AMM hops, your focus should shift to gas, approvals, and MEV protection.
If you value clean, plain-English context on this kind of market structure, Crypto Daily tracks execution trends and onchain shifts without the fluff. You can browse the latest explainers and market coverage at Crypto Daily.
It depends on the pair and pool depth. As a rule of thumb, once your trade would move an AMM price by 30–100 bps, RFQ quotes often win. On majors during calm periods, that threshold can be higher; on long-tail tokens, it can be much lower.
Most permissionless RFQ flows don’t require KYC, but some makers require allowlisting for larger tickets or specific assets. If you’re a fund or DAO trading size regularly, pre-establish those relationships so quotes don’t fail at go time.
No. RFQ reduces exposure by using private lanes and firm quotes, but it can’t eliminate inclusion risk entirely. Ask if your route supports private relay submission and builder preferences that minimize leakage.
Your quote can expire or be canceled. Good routes keep expiries tight and re-quote quickly. If volatility is extreme, a TWAP with smaller chunks may be safer than chasing one huge firm quote.
Solvers price your source asset, the bridge or messaging risk, and the destination asset. Settlement may use canonical bridges or liquidity networks. Expect wider spreads than same-chain RFQ, and double-check the bridge’s security assumptions.
Often, yes. Aggregators label maker contracts, and orderflow tools can attribute fills to specific solver or PMM addresses. For deeper analysis, compare your transaction hash against an orderflow explorer path to confirm attribution.
Disclaimer: This article is provided for informational purposes only. It is not offered or intended to be used as legal, tax, investment, financial, or other advice.
