Ratings · Measured, not sponsored
Decentralised exchanges
Every comparison in this category quotes the fee tier a venue advertises. We did the division nobody does — thirty days of fees over thirty days of volume — and found a thirteen-fold spread between venues whose published tier sheets look broadly the same.
0.0506% realised, and 46× capital turnover. The catch is that it exists on exactly one chain, operated by one company.
Core contracts never drained in eight years, deepest liquidity on 48 chains. You pay 3.5× Aerodrome for it.
Among the most audited codebases in DeFi. Lost $128m to rounding error in November 2025; about $27m came back.
What we did · August 27, 2026We divided thirty days of protocol fees by thirty days of volume for every venue, which is the closest thing to what traders actually paid — and it disagrees with the advertised fee tiers everywhere. Then we set each venue’s incident record against the question that decides whether an exploit matters: did anyone get their money back?
The fee tier is not the fee
Thirty days of protocol fees divided by thirty days of volume — what traders were really charged, once you account for which pools they actually traded through.
Thirteen times the cost between the top and the bottom of that list, from venues whose advertised tier sheets look broadly alike. Your rate is set by the pool your trade lands in, which you cannot see before you trade and nobody reports afterwards.
The most audited contracts in defi lost $128 million
On 3 November 2025, an attacker pushed token balances in Balancer V2’s Composable Stable Pools to rounding boundaries, batched the swaps, and accumulated the precision loss into roughly $128 million. About $19.3m of osETH was clawed back by StakeWise and roughly $8m more by white-hats. The rest did not come back.
Balancer’s TVL is now $65m — under 2% of Uniswap’s. This is why the audit count is recorded on these pages but never scored on its own, and why 30% of the score is the incident record and what happened to the people who lost money.
The ranking
Click any row for the quick read, or open a company for its full profile and per-criterion scores. Measured August 27, 2026. How we rate.
HOW WE SCORED THIS — CRITERIA AND WEIGHTS+
Every DEX comparison in existence quotes the fee tier — “0.30% versus 0.05%” — and every one of them is quoting a number almost nobody actually pays. We computed what traders really paid: thirty days of protocol fees divided by thirty days of volume, per venue. The spread between the advertised tier and the realised rate is the story, and it runs from 0.05% to 0.66%.
Contract exploits with dates, amounts and — the question every other comparison skips — whether users were actually made whole. A DEX is a set of contracts holding pooled money, so this is the first question, not the fifth. Audit count is recorded but never scored on its own: Balancer V2 was among the most audited contract systems in DeFi and lost $128m to a rounding error in November 2025.
Computed by us, not read off a fee-tier page: thirty days of protocol fees divided by thirty days of volume, per venue, from public data. That single division exposes what routing, tier mix and pool design really cost, and it disagrees with the advertised tiers everywhere. It is not the whole cost of a trade — slippage and gas sit on top — but it is the part every comparison misstates.
Total value locked and thirty-day volume, both from public data at a stated timestamp. Depth is what decides slippage on a real trade, and volume is what proves the depth is usable rather than parked.
The number of chains where the venue is actually deployed, weighted by whether volume genuinely flows there. A deployment on forty chains that trades on two is coverage on paper.
Thirty-day volume divided by total value locked — how many times each dollar of liquidity turned over. Nobody publishes this, and it separates a venue whose pools are genuinely used from one sitting on idle incentive-farmed capital. It also predicts fee income per dollar of risk taken by liquidity providers.
Weights sum to 100. If we cannot verify a criterion, we delete it rather than score it on impressions — read the methodology.
Is there a free alternative?
On a small trade, the fee tier is not what costs you money. Slippage and gas are.
The realised rates on this page run from 0.05% to 0.66% — on a $500 swap that is a spread of about three dollars. Meanwhile a badly routed trade into a thin pool can cost twenty times that in slippage, and on Ethereum mainnet the gas can cost more than both combined. So use an aggregator to route the trade, check the price impact before you sign, and only then care which venue you landed on. And never grant an infinite token approval to a router you have not used before: the two largest aggregator losses on record, LI.FI and Socket, were both infinite approvals drained through a newly added route.
What changed since last time
- 2026-08-27Category published.Every comparison quotes fee tiers. We computed the rate traders actually paid, and the two numbers are not the same at any venue on this page.
Questions
Which DEX is actually the cheapest?+
On our measurement, Aerodrome — 0.0506% of volume paid in fees over thirty days, against 0.0651% at PancakeSwap, 0.0659% at Orca, 0.1762% at Uniswap, 0.2340% at Raydium, 0.3506% at Meteora and 0.6639% at SushiSwap. Note what that list is not: it is not the fee tiers those venues advertise. It is thirty days of protocol fees divided by thirty days of volume, which is the closest public proxy for what traders were really charged once the mix of pools they traded through is accounted for.
Why does the realised rate differ from the advertised fee tier?+
Because a venue is not one fee. It is hundreds of pools across several tiers, and your rate is set by which pool your trade routes through — something you cannot see before you trade and nobody reports afterwards. A venue whose volume happens to sit in 0.30% pools charges roughly six times one whose volume sits in 0.05% pools, and both can publish identical tier sheets. The mix is the price.
Do audits tell you a DEX is safe?+
They do not, and this category has the cleanest counter-example in DeFi. Balancer V2 was among the most heavily audited contract systems ever deployed, and on 3 November 2025 an attacker accumulated rounding error in its Composable Stable Pools into roughly $128 million. About $27 million came back. Curve’s 2023 loss came from a bug in the Vyper compiler rather than in Curve’s own logic — a class of failure no audit of the protocol would have caught. We record audit counts and score incidents, and we never treat the first as evidence about the second.
Is a DEX safer than a centralised exchange?+
It removes one risk and adds another. Nobody can freeze your withdrawal or lend out your balance, because you never handed over custody — that is real, and it is why the category exists. In exchange, you take on smart-contract risk with no deposit insurance and no support desk, and you take on the risk of your own mistakes: a wrong token contract, an infinite approval to a malicious router, a signature you did not read. Empirically, more retail money has been lost to approvals and phishing than to DEX contract failures.
What does capital turnover tell me as a trader?+
It is volume divided by TVL — how many times each dollar of liquidity was actually traded through in a month — and it is the fastest way to tell working liquidity from parked liquidity. Aerodrome turns its capital over 46 times a month, Uniswap 14, Curve 2. High turnover means the pools you are routing into are genuinely active; very low turnover usually means the TVL headline is being propped up by incentives rather than by trade.
How this is funded
It is not. There are no affiliate links on this page or anywhere on this site, no paid placements, and no sponsored positions. Nobody in this table can buy a place in it, accelerate their inclusion, or influence a score — and none of them paid us anything, because there is nothing here to buy. The full policy.