How to compare the full cost of a crypto swap

Compare crypto swaps by the net amount delivered for the same input, asset pair, network and timing. Then add any cost paid separately by the sender. A low advertised fee can coexist with a less favorable rate, and a larger displayed output can shrink when payout deductions are applied.

The useful comparison is a complete route from your starting balance to the usable asset at the destination. It should also show what remains uncertain. A floating estimate is not a fixed commitment, and a quote that cannot be used under your eligibility or verification conditions is not a practical bargain.

Define what you are comparing

Choose the input asset, input amount, output asset and both networks before collecting quotes. Record the rate mode and the observation time. Without those fields, two numbers may describe different transactions even when the tickers look identical.

The destination matters too. An output on one network may not be accepted by the wallet or account where you need it. A lower-cost route that requires a second transfer, bridge or conversion can introduce another fee and another risk. Compare the route that actually reaches your intended destination.

Use a consistent accounting boundary. If you compare what leaves your wallet with what arrives in another wallet, include any sending-network fee paid on top of the input. If you compare only the provider's quote, say that the separate sending fee is outside that comparison. Both measurements can be useful when their boundaries are explicit.

Do not convert unlike assets into a percentage without stating a reference value. A difference between one unit of an asset and another unit of a different asset is not automatically a fee. You need a contemporaneous reference rate or a direct comparison of equivalent quotes to interpret the difference.

Separate the rate from the explicit fee

The quoted exchange rate determines how much output the input buys before any additional deductions identified by the provider. A service can include compensation in that rate rather than display it as a separate charge. The difference from a chosen reference price is commonly discussed as a spread.

A reference price is itself a choice. It can come from a different venue, trade size or moment. Use it to explain your calculation, not to suggest that every difference is a hidden fee. Price movement, available liquidity and the mechanics of the route can all affect the result.

An explicit service fee is a disclosed charge. A payout fee is a deduction associated with sending the output. A network fee pays for activity on the relevant network. The provider's labels may combine some of these items, so read whether a displayed amount already includes them before adding anything again.

Double counting produces a misleading comparison in the other direction. If the quote already states the final received amount after a payout deduction, subtracting the deduction once more makes that route look more expensive than the displayed terms indicate. Record “included”, “additional” or “unclear” for each component.

Network costs are not one universal line item

A sending wallet may charge a network fee separately from the amount deposited. The provider may also pay a network cost when delivering the output and reflect it in the rate or a separate deduction. These are different legs, not necessarily duplicate charges.

On Ethereum, gas measures execution work and fees are paid in ETH. The official gas documentation explains the mechanism. Other networks use different models, so do not apply an Ethereum estimate to a Bitcoin, Tron or Solana transfer merely because the output asset has the same ticker.

A DEX route can involve an approval transaction and a swap transaction. A cross-network route can involve still more steps. Determine which actions are necessary for the specific transaction and which amounts are paid in a native network asset. A quote in one token may not include every balance needed to execute it.

Failed execution can also have a cost, depending on how the network processed the transaction. An included transaction that reverts is different from a transaction rejected before execution. Avoid the blanket assumption that “no swap completed” means “no network cost occurred”.

Work through a complete hypothetical example

Assume an input worth 1,000 accounting units and a reference conversion of one output unit for each input unit. These invented numbers explain the method; they are not live prices, a token peg or a quote from a real service. The sender also pays two units separately in network costs.

Route A displays 990 output units as the final amount after its deductions. Route B displays 995 before an eight-unit payout deduction. Route B therefore delivers 987. On the quoted input alone, Route A delivers three units more, despite the larger headline number on Route B.

The total starting outlay is 1,002 units in this simplified example. Route A's shortfall from that outlay is 12 units, or about 1.20%. Route B's shortfall is 15 units, or about 1.50%. These percentages depend on the stated one-to-one reference and the decision to include the separate sending cost.

If the sending cost were already included in the quoted 1,000, adding two again would be wrong. If the assets had different values, the one-to-one arithmetic would also be wrong. The example works because its accounting assumptions are explicit, not because a universal formula makes every quote directly comparable.

Why the cheapest route can change with amount

Consider another hypothetical comparison. Route A delivers 0.992 times the input. Route B delivers 0.997 times the input minus eight output units. The first route has a larger proportional reduction; the second has a smaller proportional reduction plus a fixed deduction.

For an input of 1,000, Route A delivers 992 and Route B delivers 989. For an input of 5,000, Route A delivers 4,960 and Route B delivers 4,977. The route with the fixed deduction is less attractive at the smaller amount but more attractive at the larger amount under these assumptions.

Setting the outputs equal gives 0.992 times the input equals 0.997 times the input minus eight. The difference is 0.005 times the input, so the hypothetical break-even input is 1,600. This is an explanation of fee structure, not a recommendation to increase a transaction's size.

Real quotes can change nonlinearly with liquidity, limits and rate conditions. Do not extrapolate a small quote into a large one. Sample the amount relevant to your actual decision and record what the provider displays for that amount.

Keep price impact and slippage distinct

Price impact concerns how a trade changes the price available through the liquidity it uses. A larger trade relative to available liquidity can receive a different average execution price. That is not the same as a price moving between the time you inspect a quote and the time the transaction executes.

Slippage describes a difference between an expected price or output and the executed result. A slippage-tolerance setting can define an acceptable execution boundary in a particular interface, but it does not create a universal promise about total transaction costs or success.

Read the actual minimum-received amount and the transaction conditions. A percentage setting without the corresponding asset amount can be hard to evaluate. Also distinguish a transaction that fails because its conditions cannot be met from one that executes at a less favorable amount within the permitted range.

For comparisons, record whether each output is an estimate, a minimum under specified conditions or a fixed quote with an expiry. Calling all three simply “you receive” conceals differences that can be more important than the apparent price advantage.

Quote timing can reverse an apparent result

Collect comparable quotes close together and record timestamps. If one quote was observed before a market move and another afterward, the difference does not isolate provider cost. A static screenshot should preserve the observation context rather than imply a price that remains available.

A fixed quote may require the deposit to arrive or satisfy specified conditions before a deadline. A floating quote may adjust according to the provider's stated execution process. Neither label alone tells you the relevant event that determines the final rate.

Check the deadline, required amount, supported network and treatment of late or incorrect deposits. A quote that expires before your transfer can meet its conditions may not be usable. The lowest displayed number for cost is therefore only one input into the decision.

Do not repeatedly send test deposits merely to compare advertised prices. Quote sampling and funded execution are separate forms of evidence. A quote can be inspected without pretending that the settlement process has been tested.

Build a comparison you can explain

For each candidate, record input amount, output amount, rate mode, quote time, network, stated service charge, payout deduction and separate sender cost. Mark which items are already included. If a material deduction is unclear, write that uncertainty into the comparison instead of treating it as zero.

Then record the custody model, later-verification policy and refusal or refund conditions. These fields do not belong inside the arithmetic, but they determine whether the quoted route fits the reader's requirements. A financial comparison that ignores an unacceptable access condition is incomplete.

Compare net outputs first under a shared boundary. Use percentages only when the reference value and denominator are stated. If the quote is floating, describe the result as an observation rather than a future payout. If a fee can vary, identify it rather than silently substituting a precise estimate.

The Juno Swap methodology keeps cost transparency separate from policy and risk warnings. Read fixed versus floating rates to interpret the commitment behind a quote, and network selection before deciding that two outputs are interchangeable.

Questions readers ask

Is a zero service fee the cheapest option?

Not necessarily. Compensation can be reflected in the rate, and network or payout deductions can change the delivered amount. Compare the net output and separately paid costs under the same assumptions.

Should I always choose the largest displayed output?

Only after establishing whether it is final, estimated or before deductions. The asset, network, rate mode and timing must also match. An incompatible destination makes a larger number irrelevant to the intended transfer.

Can I use one quote to rank every transaction size?

No. Fixed deductions, proportional charges and liquidity effects can change the result with amount. The hypothetical break-even example illustrates why a route can be more competitive at one size and less competitive at another.

Is the difference from a market price entirely a fee?

No. It depends on the reference price, timing, size and execution conditions. Explain those assumptions and distinguish explicit charges from the observed rate difference rather than assigning an unsupported cause to the entire gap.