Uniswap

After Siemens laid him off on July 6, 2017, Hayden Adams learned Ethereum development and chose to implement an automated market maker described in posts by Vitalik Buterin. His own history draws a careful line around authorship: Alan Lu first conceived x*y=k market makers on Ethereum, Buterin publicized the design, and Adams created the specific implementation that became Uniswap.

That distinction does not diminish the work. Adams turned a mechanism into a production system, secured a documented $100,000 Ethereum Foundation grant, worked with Runtime Verification, and launched v1 on November 2, 2018. About $30,000 from one provider, split across three tokens, supplied the first liquidity. Uniswap's achievement was implementation, accessibility, and persistence—not ownership of the AMM idea.

V1's constraint was visible in every pool: an ERC-20 traded through ETH. V2, launched May 18, 2020, allowed direct ERC-20 pairs and added oracles and flash swaps. V3 reached mainnet on May 5, 2021 and let LPs concentrate capital inside chosen ranges, exchanging passive simplicity for active position risk.

V4, live from January 31, 2025, changed the boundary between core and extension. A singleton PoolManager and flash accounting form a non-upgradeable settlement layer, while optional hooks can run custom code around swaps and liquidity changes. The old protocols did not disappear. They remain callable, so 'Uniswap' is a family of contracts with different liquidity, fees, licenses, and risks rather than software governed by one upgrade switch.

Hooks make the family more expressive and less uniform. A hook can alter math, call an oracle, depend on a bridge, hold assets, or retain upgrade keys. The Foundation's own security framework says it does not certify hook implementations. The pink interface may route across pools, but its visual continuity is not a security guarantee.

UNI launched on September 16, 2020, nearly two years after v1. One billion tokens were minted at genesis: 60% for community members, 21.266% for team and future employees, 18.044% for investors, and 0.69% for advisers. The initial allocation described inflation beginning after four years. The published Uni.sol contract permits the authorized minter to mint at most 2% of the contract’s totalSupply per call, with at least 365 days between mint calls. This is a callable authority, not automatic yearly issuance, and its denominator is contract total supply rather than circulating supply.

What UNI controls is concrete but bounded. Delegates can vote to spend treasury assets, set protocol fees, and use the capped mint power. They cannot alter the code of already deployed core contracts. Holding UNI alone does not grant Labs shares, an LP position, redemption at a fixed value, or an automatic slice of each swap fee.

Practical control also depends on participation. Voting power must be delegated, proposals must clear process and quorum, and many users of the protocol never vote. The token connects people to specific governance levers; it does not collapse all developers, LPs, interfaces, and chains into one owner.

Unichain mainnet opened on February 11, 2025 as an OP Stack L2 where users could access Uniswap v2, v3, and v4. This put sequencing, bridging, fault proofs, and L2 governance beside pool-level risks. Even when a swap uses Uniswap contracts, its execution now inherits assumptions from the chain carrying it.

DUNI addressed a different gap. Adopted as a Wyoming decentralized unincorporated nonprofit association, it lets Uniswap Governance sign contracts, hire service providers, and handle legal and tax duties. Membership is fluid: a person must hold more than zero UNI and participate by delegating, voting, or proposing. The arrangement explicitly left the token, protocol, and core voting mechanics unchanged.

The legal wrapper also made formal division of labor possible. DUNI can bind service providers; the Foundation and administrators carry limited ministerial duties; Labs builds products and protocol software as a separate company. Those links create accountability paths, but they do not make every Labs roadmap choice a direct token-holder command.

UNIfication passed onchain and executed on December 28, 2025. Its eight actions included sending 100 million treasury UNI to 0xdead, turning on v2 and selected v3 fees on Ethereum, approving up to 40 million UNI to vest to Labs over two years, and anchoring a DUNI service agreement. The same vote therefore reduced supply and authorized substantial treasury spending. Sending UNI to 0xdead removes it from economic use but does not itself decrement the token contract’s totalSupply.

The recurring burn is an exchange, not a payout. Fee adapters move assorted assets into an immutable TokenJar. When that basket is worth enough, a caller burns the configured UNI threshold through a releaser such as Firepit and withdraws the basket. The caller captures the spread; all UNI holders experience the supply reduction, but they do not receive the collected tokens pro rata.

On July 27, 2026, proposal 100 extended fee control to specified v4 pool families on seven chains. Governance may replace the fee policy or override pool and family settings even though v4 core remains immutable. UNI's economic link to usage is now real, but indirect and governed: its result depends on trading volume, fee choices, LP response, searcher participation, cross-chain delivery, future minting, and treasury outflows.