Proposal 100 takes roughly one-sixth of every swap fee and routes it into UNI burns. Liquidity providers keep their yields intact.
Uniswap activated protocol fees across seven v4 deployments on 27 July after Governance Proposal 100 cleared quorum with more than 6 million UNI to spare. Approximately 46.6 million UNI voted yes; 1.27 million voted no. The required threshold was 40 million.
Formal on-chain voting ran from 19 to 26 July, following a temperature check that opened on 7 July. Both stages produced lopsided results in favour of the switch — the sort of margin that has been rare in Uniswap governance since the DAO's founding, and near-impossible on any question involving the flow of fee revenue.
The mechanic is deliberately restrained. The protocol takes roughly one-sixth of the existing swap fee: on a standard 30 basis point pool, five basis points route to the protocol contract on top of the pool fee, and liquidity providers keep their existing yields. Founder Hayden Adams was explicit throughout the temperature-check thread that the design was engineered not to disturb LP economics, on the theory that liquidity is the moat and cannibalising it to feed token holders would be self-defeating.
Fees do not go to UNI holders directly. They flow into TokenJar contracts that require burning UNI to claim, turning revenue into permanent supply reduction rather than a dividend. That distinction matters legally as much as economically. Direct payments to token holders raise obvious securities concerns; a burn mechanism reduces supply without a direct payment, which is closer to a buyback structure. The Wells notice served to Uniswap Labs in April 2024 shaped how any distributive mechanism could be structured. Proposal 100 reads like the answer legal spent two years drafting.
The seven networks covered are Ethereum, Arbitrum, Base, BNB Chain, Polygon, OP Mainnet, and Robinhood Chain — the last a permissionless Layer 2 that Robinhood launched on 1 July, with Uniswap as its day-one AMM. That inclusion is not incidental. Every dollar of swap volume Robinhood's chain routes into Uniswap now sends a slice through the burn engine.
Early revenue is running at roughly $325,000 a day. Annualised without any growth assumption, that puts protocol take in the low nine figures. Modest against the DEX's monthly volume, and modest against UNI's market capitalisation, but a real number where the previous number was zero.
Proposal 100 builds on the UNIfication framework the DAO approved in late 2025, which enabled protocol fees and burns on v2 and select v3 pools as a proof-of-concept. What was small-scale and cautious there is now default and comprehensive here. Uniswap Labs will still control the fee toggle on individual v4 pools; governance now controls whether the mechanism is on at all.
The pool of DeFi tokens that reliably return revenue to holders is small. Aave and MakerDAO have been in it for years. Uniswap, despite dominating DEX volume for most of its existence, has resisted joining until now. The counter-argument has always been that turning on fees would push volume to competitors like PancakeSwap, 1inch, or Aave's V4 unified liquidity layer — competitors that don't skim. The 46-to-1 margin says the DAO has stopped worrying about that.
Second-order effects will surface quickly. LPs migrating between pools now have to consider whether the pool they're leaving does or does not send fees to burns. Wallets and aggregators will surface effective yields differently once the deduction is standard. Analysts modelling UNI's value have a real revenue stream to plug in for the first time since Uniswap became the dominant DEX in 2020.
For years the answer to what UNI actually does was that it lets you vote on things Uniswap Labs mostly does anyway. As of 27 July, it also captures a fraction of every trade the protocol clears.
