Flash Loan Fee

Different protocols have different business models. For a lending market like Aave, the flash loan fee is direct revenue: it accrues to the protocol treasury and is shared with the lenders who supplied the underlying asset, compensating them for the small operational risk of being part of a flash loan cycle. For a DEX or a protocol whose primary product is not lending — dYdX, Balancer, Morpho Blue, Euler — the flash loan is a value-added feature, not a revenue line. The protocol benefits from the extra transaction volume, from integrating with arbitrage bots that keep its prices efficient, and from being the venue of choice for sophisticated DeFi users. Offering flash loans for free costs them nothing (since the loan is atomic and carries no risk) and earns them goodwill and volume. The result is that the same asset can be borrowed for 0 bps on one venue and 5 bps on another, and the choice between them comes down to liquidity depth and asset coverage rather than cost.

The break-even spread is the minimum gross return your strategy must generate in order to cover the flash loan fee. It is numerically equal to the fee rate, because the fee is charged as a percentage of the loan. If the fee rate is 5 bps, your arbitrage must produce at least 0.05% gross return on the borrowed amount to break even; anything less is a loss. This makes the fee rate a direct gate on strategy viability: opportunities with spreads below 5 bps are only viable on zero-fee venues, and opportunities with spreads above 30 bps are rare enough that most flash loan strategies target the 5–15 bps range where Aave's 5 bps fee still leaves meaningful profit. The calculator's advanced panel shows this break-even explicitly, along with the net profit for any assumed gross spread, so you can quickly evaluate whether a specific opportunity is worth executing.

Technically, a flash loan can be as large as the protocol's available liquidity in the asset you want to borrow. Practically, the limiting factor is almost never the protocol — it is the depth of the market you are trading into. If you borrow $100M to arbitrage a DEX pool with only $20M of liquidity, your own trade will move the price against you, and the arbitrage will produce far less than expected. For most strategies the optimal loan size is the largest amount that can be traded without significantly moving the price on both legs of the trade — typically a few percent of the smaller pool's reserves. Zero-fee providers tend to have less liquidity than Aave, so a very large loan may only be possible on a paid venue, which changes the cost calculus. This calculator lets you price both scenarios side by side so you can see exactly how the trade-off plays out at your intended size.

The transaction reverts. This is the defining safety property of flash loans: because the repayment must complete within the same atomic transaction, any strategy that fails to produce enough to cover the principal plus fee simply does not execute. You cannot end up in a position where you owe money you do not have — the worst outcome is a reverted transaction that costs you the gas. In practice this means flash loan strategies are "riskless" in the sense that a bad trade cannot lose you the borrowed principal. The real risks are subtler: gas costs on a reverted attempt, MEV extraction by a searcher who front-runs your transaction, slippage that eats into the profit margin, and smart-contract risk in the protocol or in your own code. On a well-designed strategy these are small relative to the potential profit, which is why flash loans are such a popular tool for sophisticated DeFi operators.

Use the free provider whenever it works. If the asset you need is available on dYdX, Balancer, Morpho Blue, Euler or MakerDAO in sufficient size, there is no reason to pay 5 bps to Aave. The only reasons to choose Aave are: (1) the asset is not available on a free venue; (2) the free venue does not have enough liquidity for your target loan size; (3) your strategy requires a specific integration pattern that Aave's flashLoan interface supports more cleanly. For most common arbitrage on ETH, USDC, DAI, WBTC and other major assets, a free venue is the default, and Aave's 5 bps is the fallback when the free venues cannot accommodate the trade. The savings are meaningful at scale — on a $10M loan the difference between 0 and 5 bps is $5,000 per execution, which over hundreds of executions per year is a real number.

Because a Uniswap flash swap is not a flash loan in the lending sense — it is a swap that is settled at the end of the transaction instead of at the beginning. When you do a flash swap, you receive the output token first, use it for whatever you want, and then either pay back the input token plus the swap fee or return the output token. The fee you pay is the pool's normal swap fee (0.05%, 0.30% or 1.00% depending on the tier), not a separate flash-loan fee. In practice this makes flash swaps more expensive than true flash loans for most arbitrage, but they have a specific advantage: they let you borrow from a pool's reserves without needing a separate lending protocol, and they can be used atomically in complex swap paths where a standard flash loan would require multiple protocol integrations. For tight-spread arbitrage on liquid pairs, Aave at 5 bps or a free venue at 0 bps is almost always cheaper than Uniswap's 30 bps minimum on V2 or the 5 bps floor on V3's cheapest tier.

No — the fee is only the direct cost charged by the lending protocol. The full economic cost of a flash loan strategy includes several other components. Gas: executing a flash loan transaction on Ethereum mainnet costs anywhere from $10 to $500 depending on network congestion and the complexity of the strategy. Slippage: the difference between the expected price and the price you actually get on the trades that constitute your arbitrage, which can be much larger than the flash loan fee on illiquid venues. MEV: if your transaction is visible in the mempool, searchers may front-run it, capturing some or all of the profit before your transaction lands. Failed-attempt gas: even reverted transactions cost gas, and if you try several strategies that do not pan out, those costs accumulate. On a well-run strategy the total of these costs is usually dominated by gas and slippage, not the flash loan fee itself — but the fee is still the hard floor below which the strategy cannot be profitable, which is why getting it right matters.