For one route, 0.2 ETH to USDT on BNB Smart Chain, quotes taken between 05:11 and 05:19 UTC on 17 October 2026 showed Rango's top route (Mayan Aggregator) at 484.74 to 484.88 USDT out, and LI.FI's AcrossV4 route at 484.51 to 484.63 USDT out. That is a 0.03 to 0.06 percent output difference, on a $485 trade. The route that quoted the higher output also quoted a slower speed, 45 seconds against an estimated 2 seconds, so a rate check and a speed check really can come back with two different winners.
Which platform gives the best rate for ETH to USDT today?
In this sample, the Rango-surfaced route quoted the higher executable output. After normalizing each quote against its own platform's ETH oracle price, the Rango route landed at 0.99905 of mid and the LI.FI route at 0.99874 of mid, about 0.031 percent apart. That is small enough that it can flip between one quote and the next, and it is one asset pair, one amount, on one morning. The practical answer to the query is the process: read the final output after fees on the route list, then check the speed estimate on the same screen. Both aggregators let you do that before you sign anything, and the numbers move as the market moves.
What is the difference between the market exchange rate and the rate you actually get?
The market exchange rate is the mid-price. Binance spot ETHUSDT showed 2429.31 bid and 2429.32 ask at 05:12:26 UTC that morning. CoinGecko put ETH at $2426.56 with USDT at $0.998924 at about the same minute. Rango's own price feed inside the routing response said ETH $2426.07 and USDT $0.998936.
The executable output is different from every one of those numbers. It internalizes liquidity pool fees, the bridge leg, the network fee, and the aggregator's own fee line. Quoted executable rates came in at 2422.55 to 2424.41 USDT per ETH across all seven quotes, which is 0.17 to 0.31 percent below the mid that each platform's own oracle implied. That spread is the real cost of executing the route, and it is the number that matters, not the chart price.
Rates: what 0.2 ETH actually quoted
Request amount in every case: 0.2 ETH, roughly $485 at those prices. Destination asset in the cross-chain samples: USDT at address 0x55d398326f99059fF775485246999027B3197955 on BNB Smart Chain, so both aggregators were priced into the same token contract.
-
Rango dApp, route: Mayan Aggregator. 05:11:16 UTC, request id 1d363028-547f-4897-ac52-b638f5f57c77: output 484.881148 USDT, network fee 0.000041096 ETH, "Rango Fee" line 0.0003 ETH marked as taken from the output. Quote round trip measured at 1328 ms.
-
Rango dApp, same route, 05:12:47 UTC, request id 04058b2b-90b4-4957-aa9d-89adae3e0800: output 484.738098 USDT. Quote round trip 1482 ms.
-
Rango dApp, same route, 05:13:38 UTC, request id 6763aae7-7525-425d-8703-95161b0f5766: output 484.784906 USDT, oracle shift to $2426.23. Quote round trip 1419 ms.
-
LI.FI quote API, 05:18:59 UTC, id a255b754-e555-4e57-84dc-3bdd432cccb0:0, tool AcrossV4: output 484.586540 USDT, executionDuration estimate 2 s.
-
LI.FI repeat, 05:19:16 UTC: output 484.509059 USDT. LI.FI repeat, 05:19:25 UTC: output 484.625986 USDT.
Three of the Rango quotes and three of the LI.FI quotes still varied internally by up to 0.05 percent inside each platform, in minutes, with the whole market barely moved. That drift alone is enough to swallow a 0.03 percent platform difference.
Speed: how fast did each route say it would be?
-
Rango-surfaced Mayan route: estimated time 45 s, plus a recorded history stat on the route of min 12 s, avg 31 s, max 71 s.
-
LI.FI-surfaced AcrossV4 route: execution estimated at 2 s.
-
A same-chain control sample quoted on the Rango dApp at the same time, 0.2 ETH to USDC on Ethereum via an OKX DEX Aggregator route, showed an estimated 45 s with a history stat of min 13 s, avg 33 s, max 158 s. The cross-chain route out-quoted the same-chain route on speed in this one sample, which is a good reminder that same-chain does not automatically mean faster routing.
Quote latency, meaning the time to receive the routing answer, not the time to settle: Rango routes answered in 1419 to 1482 ms when requested from inside the open dApp, the same-chain route in 739 ms, and the LI.FI quote API answered in 1.52 to 1.65 s on three direct requests.
Chain-level context matters for reading those headline estimates. Chainlink's CCIP finality table lists Ethereum at roughly 12.8 minutes to hard finality, and BNB Chain at 5 seconds. BNB Chain's own documentation says the chain finalizes within two blocks in most cases, and Ethereum blocks arrive about every 12 seconds. A route cannot settle a destination payout faster than the destination chain lets it be real, and 2 s is an estimate for the bridge leg, not a promise about a complete journey, which still includes a source chain transaction, a wallet signature, and destination settlement.
Does the fastest cross-chain route also give the worst rate?
In this sample, yes on the quote screen: the 2 second Across route priced at the bottom of the range, 484.51 to 484.63 USDT, and the 45 second Mayan route at the top, 484.74 to 484.88 USDT. Fee lines as reported point the same way.
-
Rango-surfaced route, reported cost from source wallet: network fee 0.000041 to 0.000044 ETH, about $0.10, plus the 0.0003 ETH "Rango Fee" line, about $0.73 at those oracle prices. Reported total, about $0.83, roughly 0.17 percent of the notional.
-
LI.FI-surfaced route: gas estimated at 0.0000415 ETH, $0.1007, a "LIFI Fixed Fee" of 0.0005 ETH, $1.2130, plus relayer fee lines under $0.06. Reported total, about $1.36, roughly 0.28 percent.
Two honesty notes here. First, the two APIs place their fee lines in different places and neither swap was executed, so it cannot be proven from a transaction which fee lines were already subtracted inside the quoted final output; both totals above are exactly what each API self-reported next to its price. Second, this is one pair at one amount. On other pairs, or at 20 ETH instead of 0.2, a different swapper can come out ahead on either number.
Why is a 45 second cross-chain quote possible when Ethereum finality is 12.8 minutes?
Because these routes do not necessarily wait for the source chain's hard finality before paying out on the destination. They rely on pre-confirmation style mechanics, auction or relayer based, where the solver pays out on the destination chain before the source chain reaches full finality. The Mayan technical FAQ, the swapper inside the Rango route quoted here, describes what happens when price moves while a swap is in flight and how a minimum output guard works, rather than a finality wait. That mechanism is what makes a 45 s estimate mathematically possible against a 12.8 minute finality background, and it is also why the FAQ talks about guarantees in terms of minimum output rather than exact output.
How to quote ETH to USDT yourself before you sign
-
Open the route directly, for example the Rango ETH to USDT page, pick your amount, and let the route list load.
-
Read the final output after fees, not the raw market rate. If the screen shows a time estimate and a historical route stat, read both, the Rango walk-through of its swap flow describes checking route details, amount and fees before confirmation.
-
Compare at least two aggregators against the same destination token contract, as this test did against LI.FI, so the comparison is of routes, not of differently boxed assets.
-
Note the platform's own oracle price against an independent reference like Binance spot or CoinGecko, so you can tell the difference between a bad route and a moving market.
What this test can and cannot tell you
These were quote screen records, not executed swaps. One pair, one 0.2 ETH size, seven quotes over nine minutes on 17 October 2026. Both platforms' price feeds, and Binance and CoinGecko reference prices, sat within about $3 of each other during the window. The result says the two aggregators priced within roughly 0.06 percent of each other on this route while quoting very different speeds. It does not say either platform is always cheapest, always fastest, or correct about chain counts, and the cheapest realistic way to find out for your trade is still to read the final output and the time estimate on your own route before you confirm.
*Sources: rango.exchange/swap/eth-to-usdt (route page), rango.exchange/learn/crypto-basics/swap-crypto-rango (swap flow), docs.chain.link/ccip/concepts/execution-latency/finality-by-chain (finality table), docs.bnbchain.org/bnb-smart-chain/introduction (BSC finality), ethereum.org/developers/docs/transactions (block cadence), docs.mayan.finance/resources/technical-faq (in-flight swaps), api.binance.com ETHUSDT ticker, api.coingecko.com ETH price, li.quest/v1 quote API (LI.FI's own response), rango dApp routing responses captured 17 Oct 2026 05:11 to 05:19 UTC.*