Best DEX aggregator by quote output: who actually gives you more
Basis points of output a DEX aggregator gives up against the best quote fired at the same moment, on the same pair, from the same place.
TL;DR. As of , KyberSwap posts the lowest quote shortfall at 0.58 bps (p50, 24h) across 6 ranked quote sources. Source: OpenChainBench, https://openchainbench.com/benchmarks/evm-quote-output.
Read this carefully
Every provider is asked for the same swap at the same second, and each answer is scored against the best of that round. The figure is how much less you would have received, in basis points. Quotes are not comparable between rounds: the market moves 30 to 37 bps in five minutes while the field sits inside 25. Two providers take a fee out of the quote and declare it, shown separately: giving up 25 bps to a toll is not routing badly.
Everyone publishes how fast their quote API answers. Almost nobody publishes what the quote is worth. This page asks every aggregator for the same swap at the same second and compares what each one offers against the best of them, so the number is a difference rather than a price. The spread is smaller than the marketing suggests and larger than it looks: the leaders sit within a couple of basis points of each other, while two well-known names give up fifteen and twenty-five, and that gap turns out to be their own declared fee rather than their routing.
Methodology
The harness fires one pair a minute at every provider in parallel from three regions, keeps each successful quote, and measures it against the best quote of that same tick. Output is normalised to whole destination tokens first, because one provider returns whole tokens where the rest return the smallest unit. A tick with a single answer records nothing: zero there would hand a provider a perfect score for being the only one that replied. What this page does not yet do is subtract gas. Providers report it in four incompatible units and their USD figures disagree by a factor of 3.5 on the same pair at the same block, so a net-of-gas column would be precise and wrong. It is the next piece of work and it is named here rather than left for a reader to discover.
Results: 10 quote sources measured, 6 ranked by Quote shortfall (p50, 24h)
| № | Quote sources | Shortfall | Near-best ticks | Declared fee | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | KyberSwap | 0.58 bps | 68.1% | — | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 2 | Nordstern | 0.65 bps | 60.8% | — | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 3 | ParaSwap | 0.79 bps | 39.3% | — | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 4 | Uniswap | 2.34 bps | 16.3% | — | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 5 | Mobula | 12.5 bps | 0% | — | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 6 | 0x | 18.5 bps | 0% | 12.0 bps | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Measured, not ranked | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| not ranked | RouteRobinhood only | 0.36 bps | 69.8% | — | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| not ranked | CoW ProtocolRFQ or batch | 3.79 bps | 2.85% | — | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| not ranked | LI.FIRelayer | 24.8 bps | 0.044% | 25.0 bps | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| not ranked | BebopRFQ or batch | 162 bps | 0.30% | — | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Source code github.com/ChainBench/OpenChainBench/tree/dev/harnesses/evm-swap-quoting