Two screens can both show a “price” for a token and still describe different things. A swap interface on Solana usually returns a quote: an estimated output for one specific input amount, routed through one or more liquidity sources, with a price-impact figure and a slippage tolerance attached. A perpetual venue with an order book shows something else: resting bids and asks at discrete prices, each with a size. Both lead to an execution cost, but the numbers are produced differently, mean different things and leave different evidence behind. This article compares them with fictional numbers. It is a mechanics explainer, not a venue recommendation.
What a swap quote is answering
An aggregator takes an input token, an output token and an amount, then searches the liquidity sources it can reach for a route. Its answer is conditional: if a transaction executed now, along this route, roughly this much would come out. Many of those sources are automated pools rather than lists of resting orders. A pool’s pricing rule sets the rate from its reserves, so the rate worsens as your trade grows relative to those reserves. The StonkBuilder explainer on how liquidity pools price a swap covers that mechanism, and its overview of how Jupiter searches Solana swap routes shows where one aggregator fits.
Interfaces differ in what they display and how they label it. For a field-by-field reading of one interface, see the guide to reading a Jupiter quote before signing. The general point holds across aggregators: a quote is an estimate for one size at one moment, not a fill.
Price impact, slippage tolerance and measured slippage
Three numbers are easy to confuse, and they answer different questions.
- Price impact is the quote’s estimate of how much your own trade moves the rate against the available liquidity. It depends on size and depth, and it is known, as an estimate, before you sign.
- Slippage tolerance is a limit you choose: the worst deviation from the quoted output you are willing to accept between quote and execution. Many swap interfaces translate it into a minimum acceptable output; if execution would deliver less, the transaction is meant to fail rather than complete at a worse rate. A tolerance is not a cost estimate. Setting 1% does not mean you pay 1%.
- Measured slippage is what actually happened, compared with the quote. It is known only after execution.
The StonkBuilder guide to slippage versus price impact separates an order’s own effect from other changes between quote and execution. Here is a fictional swap of 1,000 USDC into a token called XYZ, before any route or network fees, which a real quote should itemize:
| Field (fictional) | Value | Meaning |
|---|---|---|
| Reference rate before your trade | 5.00 XYZ per USDC | 5,000 XYZ at zero size |
| Quoted output | 4,950 XYZ | Estimate including about 1.0% price impact |
| Slippage tolerance | 0.5% | Your chosen boundary |
| Implied minimum output | 4,925.25 XYZ | 4,950 × (1 − 0.005) |
| Actual output after confirmation | 4,940 XYZ | Known only afterward |
Measured against the quote, slippage was 10 XYZ, about 0.20%. Measured against the zero-size reference, the shortfall was 60 XYZ, about 1.2%. State which reference you use. On Solana, network fees are paid in SOL, and the linked quote guide notes that fees can apply even when a swap fails.
What an order book shows instead
An order book lists resting limit orders: bids to buy and asks to sell, each at a price with a size. Hyperliquid’s order-book documentation says its book works essentially like those of centralized exchanges while running on-chain, matches orders by price-time priority, and requires prices to be multiples of a tick size and sizes to be multiples of a lot size. Other venues publish their own rules.
Suppose a fictional perpetual book shows asks of 100 units at $20.01, 200 at $20.02 and 300 at $20.04, against bids of 150 at $19.99 and 250 at $19.98. The midpoint is $20.00. A 500-unit market buy against that unchanged snapshot would take 100 × $20.01 + 200 × $20.02 + 200 × $20.04 = $10,013, an average of about $20.026. That is roughly 13 basis points above the midpoint before fees. The SuperQuant method for walking a book level by level explains the arithmetic and why one snapshot cannot prove the depth will still be there when your order arrives.
Two differences stand out. The book’s cost rises in steps as levels are consumed, while a pool’s rate worsens continuously along its curve. And displayed book size can be withdrawn or added in the moment before your order lands, so depth is a snapshot, not an inventory.
The two numbers side by side
| Question | Aggregator swap quote | Perpetual order book |
|---|---|---|
| Where liquidity sits | Pool reserves and other sources along a route | Resting limit orders at discrete prices |
| How size changes price | Continuous along the pool’s curve | Stepwise, level by level |
| Pre-trade cost signal | Price impact on the quote | Spread plus the depth walk |
| Your protective setting | Slippage tolerance, relative to the quote | Limit price, an absolute level |
| Not included | Later market moves; fees shown separately | Fees, funding, later book changes |
A slippage tolerance resembles a limit price in that both bound the worst acceptable outcome. The tolerance is expressed relative to a quote that can itself change; a limit price is a fixed level that may simply not fill. A market order on a book carries no price bound of your own; check the venue’s order-type documentation for any protections it applies. The SuperQuant comparison of crypto order instructions covers those trade-offs. Size cuts both ways, too: a thin market can be cheaper to enter than to leave, as the guide to why selling can be harder than buying explains.
What each record lets you audit afterward
A completed swap leaves an on-chain transaction. Its record can show the signature, whether it succeeded or failed, the fee paid and the token balance changes, so you can compare actual output with the quote you saved and the minimum it implied. The StonkBuilder walkthrough on checking a Solana transaction signature explains how to read that status on the correct cluster. What the transaction cannot show is the quote you saw or what other routes would have returned, unless you recorded them.
A perpetual order leaves venue fill records: price, size, fee and time for each fill, with partial fills appearing as separate entries. From those you can compute an average fill and compare it with a book snapshot saved at decision time. Without that snapshot, the depth you faced cannot be reconstructed. Fees follow the venue’s current schedule and your tier; see the SuperQuant example of maker and taker fee arithmetic.
In both cases the venue does not preserve your pre-trade view. Save the quote or snapshot, the timestamp and the reference price yourself.
Where research tools stop
SuperQuant does not route swaps or place exchange orders. Its bundled histories are synthetic hourly candles, and its backtests apply simplified assumptions, by default 2 basis points of slippage and a 0.05% fee per fill, according to its research methodology. Hourly candles contain neither pool reserves nor book depth, so a backtest cannot reproduce a swap quote or a book sweep. Treat cost inputs as assumptions to stress, not as measurements of what a real order would have cost.
Keep learning
The StonkBuilder Jupiter swap basics course puts routes, wallet connection, quotes and price impact into one short sequence. For the contract side, SuperQuant’s spot versus perpetual futures guide explains what a perpetual position adds beyond execution cost. This is educational research context, not investment advice.