Fragmented liquidity is the central inefficiency of decentralised trading, and 1inch has been solving it since 2019. Its Pathfinder algorithm splits an order across pools, venues and intermediate assets to find the best net price, and the result is routinely better than trading directly against any single pool.
Fusion: the part that matters most
Standard aggregation still leaves a user exposed to sandwich attacks — a bot sees the pending transaction, front-runs it, and sells into the price it moved. Fusion replaces that with a Dutch auction: the user signs an intent, and professional resolvers compete to fill it, absorbing gas and MEV risk themselves. For ordinary users this is a genuine improvement in realised price, and it is one of the few production systems where MEV protection is the default rather than an opt-in setting nobody finds.
Costs
No protocol fee on classic swaps. Fusion orders are priced by resolvers, who take their margin in the fill rather than as a stated fee, so the comparison to make is the final amount received rather than a headline percentage. Gas on classic swaps is paid by the user; on Fusion it is embedded in the resolver's quote.
Risks
Two are worth stating. Router approvals: aggregation requires granting the 1inch router permission to move your tokens, and a broad standing approval is a liability if it is ever exploited or if you sign a malicious transaction elsewhere. Revoke what you are not using. Resolver opacity: Fusion depends on a set of professional market makers whose incentives you take on trust — they are competing, which disciplines pricing, but you cannot audit the auction directly.
Who should use 1inch
Anyone swapping on EVM chains, particularly for larger orders where routing matters and for anyone who has been sandwiched before. Compare its quote against the direct pool for small trades, where gas can outweigh routing gains, and manage your approvals actively.