Uniswap v4 has crossed 90,000 unique hooks initialized and linked to deployed pools, a fourfold increase from roughly 22,600 earlier in 2026. The milestone marks the protocol's transition from a single-product automated market maker to a programmable liquidity platform — one that now processes 48...
Uniswap v4 has crossed 90,000 unique hooks initialized and linked to deployed pools, a fourfold increase from roughly 22,600 earlier in 2026. The milestone marks the protocol's transition from a single-product automated market maker to a programmable liquidity platform — one that now processes 48% of Uniswap's weekly swap volume, up from near zero at launch on January 30. Protocol revenue since the December 2025 fee switch activation has reached $23.15 million, with $38 billion in cumulative volume flowing through hooked pools.
The growth comes with a security cost. A September 14 report from 0x Labs classified 54.2% of 84,163 analyzed hooks as malicious and another 26.4% as likely malicious, leaving just 19.4% verified safe. Two hook-related exploits — Cork Protocol ($11M, May 2025) and Bunni ($8.4M, September 2025) — have already cost users more than $20 million. Uniswap Labs maintains that its official API only routes through reviewed hooks, placing the risk on users who interact with unvetted contracts directly. The data presents a protocol at a pivotal juncture: rapid adoption of a new architecture coinciding with a surface area expansion that neither auditors nor aggregators have fully mapped.
Uniswap v4 replaced the factory model used in v2 and v3 — where each pool deployed its own contract — with a single PoolManager.sol singleton contract that holds all pool state. The architectural shift delivers measurable cost reductions. Pool creation gas costs fell by approximately 99%, according to Uniswap Labs benchmarks, because adding a pool is now a state update rather than a contract deployment. Multi-hop swaps cost roughly 50% less in gas, as intermediate token transfers between separate contracts are eliminated.
The singleton architecture works in tandem with flash accounting, enabled by EIP-1153 transient storage. Rather than settling token balances after each swap in a multi-step transaction, the system records balance changes in transient storage and nets them against each other, settling only the final delta. For a three-hop swap that previously required six token transfers, flash accounting reduces the operation to two.
Hooks are the third architectural component. Each pool can attach an external contract — the hook — that executes custom logic at defined points in a swap or liquidity event lifecycle: before and after swaps, before and after liquidity modifications, and during fee calculations. The design turns Uniswap from a protocol with a fixed execution path into a platform where third parties program the trading logic.
The adoption trajectory since v4's January 30 mainnet launch has been steep:
| Metric | Value | Period | |---|---|---| | Unique hooks initialized | 90,000+ | As of September 2026 | | Hooks earlier in 2026 | ~22,600 | January 2026 | | V4 pools tracked | 4,689 | September 2026 | | V4 weekly volume share | 48% | September 2026 | | V3 weekly volume share | 52% | September 2026 | | V4 weekly fee revenue share | 25% | September 2026 | | V3 weekly fee revenue share | 67% | September 2026 | | Cumulative v4 volume | ~$355 billion | Through June 2026 | | Volume through hooked pools | $38 billion+ | Year-to-date 2026 |
The gap between v4's 48% volume share and its 25% fee share reflects the architecture's lower cost structure — v4 pools attract volume partly because they are cheaper to trade through. Uniswap's Universal Router splits orders across v2, v3, and v4 liquidity in a single transaction, selecting whichever version offers better execution. This design explains the gradual rather than abrupt migration: v4 gains volume organically as its pools offer tighter spreads.
Chain-specific adoption varies. On Robinhood Chain, v4 accounts for 42% of weekly trading volume. On Base, 11%. On Arbitrum and Optimism, 18% each. On Circle's Arc chain, launched in September, v4 already contributes 29.7% of daily volume.
On September 10, Uniswap Labs deployed StablePair Hook, a dynamic-fee hook built for stable pairs such as USDC/USDT and WBTC/cbBTC. The hook launched on Ethereum mainnet with two initial pools: USDC/USDG and USDC/USDT.
The mechanism works by measuring how far the pool price has drifted from a reference price on every swap, then adjusting the fee proportionally. When the pool trades at peg, fees are minimal. When the price deviates, fees rise to compensate liquidity providers for impermanent loss risk. The design directly addresses a structural problem in prior versions: static fee tiers on stable pairs either overcharged during normal conditions or undercharged during volatile periods.
The context matters. Stablecoin-to-stablecoin swaps on Uniswap reached $43.4 billion in Q2 2026 — more than the next three on-chain venues combined, according to Uniswap Labs. A dynamic fee that captures more value during high-demand periods without repelling volume during normal trading represents a material improvement in LP economics for a segment that generates a disproportionate share of on-chain swap volume.
Within days of deployment, the USDC/USDT StablePair pool became the highest-volume pool on Ethereum, according to Uniswap Labs.
On September 14, 0x Labs published a report titled "Uniswap v4 hooks were a mistake." The study analyzed 84,163 hooks across six chains using static analysis, dynamic analysis, and settled-trade observations. The findings:
The primary attack vector 0x identified is "quote spoofing." A malicious hook displays an attractive exchange rate when an aggregator simulates a swap. When the user submits the actual transaction, the hook executes different logic, delivering up to 50% less than the quoted amount. Since aggregators like 0x route 70% of their $42.67 billion in 2026 volume through Uniswap liquidity, the surface area for this attack is material.
The vulnerability does not stem from Uniswap's core protocol. Both the Cork Protocol exploit ($11M) and the Bunni exploit ($8.4M) — the two largest hook-related incidents to date — originated in application-specific authorization and accounting logic built around hooks, not in PoolManager.sol itself. Cork's failure traced to missing validation in callback functions despite four separate audits, three of which excluded the hook contract from scope. Bunni's failure was a rounding bug in idle-balance accounting that an attacker exploited through 44 sequential micro-withdrawals.
Hayden Adams responded by drawing a distinction between the protocol layer and the application layer. The Uniswap API integrates only hooks that have been reviewed, meaning users interacting through official interfaces are shielded from malicious hooks. The risk concentrates on users accessing hooks through third-party aggregators or direct contract interaction.
The Trail of Bits engineering team published a security guide on July 30, identifying common vulnerability patterns in hook implementations and recommending that audits explicitly include hook contracts in scope — an omission that directly contributed to the Cork loss.
In December 2025, Uniswap governance approved the "UNIfication" proposal, activating the long-debated fee switch. The mechanism redirects approximately 17% of swap fees toward protocol revenue, which funds UNI buybacks and burns. Key revenue data since activation:
| Metric | Value | |---|---| | Cumulative protocol revenue | $23.15 million | | Daily revenue (current) | $129,274 | | 30-day revenue | ~$4.9 million | | Annualized run rate (est.) | ~$47 million | | UNI burned to date | ~$600 million equivalent proposed |
Governance votes in March and June 2026 extended the fee mechanism from Ethereum mainnet to eight additional chains, capturing revenue across the broader L2 ecosystem. The extension applied a tier-based v3 fee system to all liquidity pools by default and made protocol fee collection automatic for new v4 pools.
The fee switch transforms Uniswap from a protocol that generated zero protocol-level revenue — all fees historically went to LPs — to one with a measurable cash flow. At the current run rate, Uniswap's protocol revenue places it among the top-earning DeFi protocols, though the absolute number remains modest relative to the $355 billion in cumulative v4 volume processed.
Uniswap's total value locked across all versions stands at $3.415 billion as of mid-2026, according to DefiLlama. Monthly DEX volume reached $81.3 billion in the most recent 30-day window.
The tokenized stock category offers a window into v4's positioning for emerging asset classes. V4 holds $59.1 million in tokenized stock TVL — the largest DeFi venue for the category — followed by Kamino Lend at $41.7 million and Uniswap V3 at $20.9 million. Across V3 and V4, Uniswap added $82.8 million in tokenized stock TVL in the 30 days through mid-September 2026.
The hook architecture's flexibility positions Uniswap for use cases that rigid AMM designs cannot serve. The StablePair Hook addresses stablecoin trading. Other deployed hooks handle on-chain limit orders, dynamic fee adjustment based on volatility, and custom oracle integration. The 90,000-hook count suggests significant developer experimentation, though the 0x report's finding that only 19.4% are verified safe indicates the ecosystem is still in an early, disordered phase of maturation.
Uniswap v4 represents the largest architectural shift in DEX design since Uniswap v3 introduced concentrated liquidity in 2021. The hook system has turned the protocol into a platform: 90,000 experiments in programmable market-making, running across ten chains, processing $38 billion in volume year-to-date. The StablePair Hook demonstrates that Uniswap Labs itself is building on the platform, not just maintaining it.
The security surface area, however, has expanded proportionally. When 54.2% of deployed hooks are classified as malicious by an independent analysis — and when reviewed, audited hooks still produce $20 million in exploit losses — the system is growing faster than the safety infrastructure around it. The distinction between "protocol safe, hooks risky" is architecturally accurate but practically irrelevant to a user who loses funds through an aggregator routing to an unvetted hook.
The economic data points in one direction: v4 is becoming Uniswap. The 48% volume share, rising fee revenue, and expanding chain deployments suggest v3 will be legacy infrastructure within 12 months. The open question is whether the hook ecosystem matures into a curated, audited marketplace or remains a permissionless environment where more than half of deployed code is adversarial. The answer will determine whether programmable liquidity pools become standard DeFi infrastructure or a cautionary case study in the cost of radical openness.