Start typing to search this publication.
ONAR
Open menu

Subscribe to ONAR

Get new posts delivered straight to your inbox.

onar avatar onar
5 min read

Ask five people which cross-chain aggregator to use and you will get five brand names and zero numbers. That answer is useless. The right aggregator depends on the route you are actually swapping, and the only honest way to pick one is to compare what each aggregator's best route returns for the same trade on the same day.

That is what I did. On 12 October 2026 I pulled live quotes for several routes — ETH to SOL, BTC to SOL, and a wrapped-token variant — across the aggregators I had access to, and compared them on the four dimensions that actually matter: final output, all-in fees, speed, and how many steps the route takes. Here is what the numbers said, including where Rango won and where it did not.

What is the best cross-chain swap aggregator?

Short answer: the one that returns the most output for your specific route on the day you swap. There is no permanent winner.

The category works like this: an aggregator does not hold liquidity itself. It queries many liquidity sources — DEXs on both chains, bridges, and other infrastructure — and stitches the best combination into one execution path. Which means its value is entirely in the quality of that stitching: how many sources it queries, how it weighs fees against output, and whether the route it surfaces finishes in the asset you actually wanted.

Three aggregators showed up in my tests: Rango, which positions itself as a cross-chain swap execution aggregator covering 80+ chains; a one-step bridge product (Butter); and slower but cheap bridge-and-DEX combinations surfaced by other routers. They did not rank the same on every route. No aggregator did.

How should you actually compare aggregators? Sort by output, not by fee

This was the most expensive lesson of the test, and it cost about $11 to learn.

For the 1 ETH → SOL route, the headline numbers were:

  • Best output: Circle CCTP → Jupiter route (via Rango): 22.694 SOL, $0.25 in fees, ~26 minutes.

  • Fastest: Butter, one step: $0.18 in fees, ~3 minutes, slightly less SOL.

  • Cheapest listed fee: $0.12, ~30 seconds — and the least SOL delivered.

The route with the lowest fee is the one that returned the least. If I had sorted by "cheapest" I would have paid $0.13 less in fees and received roughly 0.1 SOL (~$11 at the time) less in output. Fee ranking is not output ranking. The all-in percentage is what you should compare, not the sticker fee.

The output gap came from where the route's cost actually lands: Ethereum-side finality, the cross-chain leg, and the Solana DEX leg each take a cut. A route can show a tiny fee and still leak value on the swap legs. Read the destination amount first, fee and time second.

Which aggregator supports Bitcoin and Solana?

Non-EVM routes are where aggregators separate themselves, because the routing problem is genuinely harder: Bitcoin has no smart contracts, so a BTC → SOL route needs a locked-asset representation somewhere in the path, and the quality question becomes what do you hold at the end?

My BTC → SOL test (0.1 BTC, quotes only, no wallet signature) made the distinction concrete:

  • Native-destination routes: every route I pulled landed native SOL on Solana. The best one returned 74.259280 SOL, $0.15 fee, ~45 seconds via the Flashnet leg.

  • The wrapped trap: starting from wrapped BTC (WBTC on Arbitrum) instead of native BTC, the cheapest quoted route still delivered native SOL — but "wrapped" describes your starting asset, and every wrap layer in a path adds a counterparty and a step. WBTC → USDC (Circle bridge) → SOL was two steps and ~26 minutes versus one step at 45 seconds.

The rule that generalizes to any aggregator: open the quote and read the destination token before you compare numbers. Wrapped is not native, and a route that ends in an IOU is not the same product as one that ends in the real asset. Rango's route display labels this; verified against the route I pulled, and I would expect the same discipline from any aggregator you use — check it, don't assume it.

One honest caveat from the same test: the worst route in the BTC → SOL sample (XO Swap) was behind by −2.39% on output. Aggregation quality is a distribution, not a guarantee. The aggregator's job is to give you the best available route; your job is to glance at the top few instead of blind-accepting the first.

How do the fee and speed dimensions actually trade off?

Across both routes, the pattern held:

Fees are also small relative to the variance: the spread between best and worst route was worth ~$11–180 depending on the pair, while the spread between best and worst quoted fee was about $0.13. When someone markets a cross-chain aggregator on fees alone, that is a category error. Rango's own learn content makes the same point from the infrastructure side — gas, transaction structure, and route count move the number that matters, which is what lands in your wallet.

So which one should you use?

The decision procedure that survived my testing:

1. Get a quote for your exact pair and amount on one major aggregator (Rango, for me, on both routes).

2. Read the destination token — native or wrapped?

3. Compare the final output against a second source's, on the same day.

4. Only then look at fee and time.

For my routes, Rango's best route won on output both times (22.694 SOL on ETH→SOL; 74.259280 SOL on BTC→SOL) while being neither the cheapest nor the fastest. If your trade is small and you value minutes over dollars, a fast one-step bridge can be the rational pick — Butter's 3-minute route lost by a hair and saved 23 minutes. If you are moving size, output dominates everything else.

None of this makes any aggregator "the best." It makes the choice legible: pick the day's best route, verify the destination asset, and stop paying $11 for $0.13 of fee savings.

Sources

Method note: all figures come from live quote responses captured on 12 Oct 2026, 06:01 UTC (ETH→SOL, requestId 03b6e8b6-e33e-47e8-b78f-91c62bb15ccc) and 13–14 Oct (BTC→SOL quote pulls, no wallet signature). Competitor figures are from those aggregators' own quote interfaces; Rango figures are from Rango's own interface.

Subscribe to ONAR