DEX trading is split beneath the market total
These are tracked on-chain DEX venues. Prediction markets, trading apps and overlapping router totals are separate. Gross trading does not tell us how many people traded or whether buyers added net money.
UNI · 19 September 2026 edition
V3 gained trading while V4 lost more. Together, the two versions handled $19.99 billion over the latest complete week, 4.5% less than the week before. A growing V3 headline does not establish stronger activity across both versions.
Completed UTC days · 12–18 Sep versus 5–11 Sep 2026 · fetched 19 Sep 2026
V3 added $0.568 billion, but V4 lost $1.520 billion: a combined decline of $0.952 billion. V4 still supplies more than half of the two-version total. Remove it and V3 alone is growing.
The longer window is stronger: together V3 and V4 handled $79.85 billion over 20 Aug–18 Sep, versus $43.01 billion over 21 Jul–19 Aug, a 85.7% rise. Weekly activity is slowing within a much stronger month.
Business reading: the weekly slowdown is concentrated in V4. A next completed seven-day total above $20.94 billion would recover the prior weekly level. More trading would strengthen the activity case, but it raises UNI’s valuation only if more value reaches holders. The published default valuation model puts UNI at $1.13 versus its frozen $9.35 market quote, about 88% lower; exchange use and token value remain separate tests.
These are gross swap volumes across V3 and V4 pools, including tokens other than UNI. The total adds each version once, excludes the Uniswap parent total and overlapping router totals, and excludes 19 Sep’s incomplete day. It is not all Uniswap versions, distinct users, net deposits or UNI-token trading volume. A routed swap can contain more than one pool leg.
Uniswap V3 daily series · Uniswap V4 daily series. Daily observations are summed into identical completed UTC windows. The market-wide readings elsewhere on this page retain their original snapshot; this UNI section is a separately dated comparison.
The largest fifteen are shown below; each share uses all 816 tracked venues, not just these rows. A busy venue may still have thin liquidity or repeated trading.
6 of six selected networks returned readings. Shares compare only these network rows and the same tracked DEX list, not the whole market. Daily change compares venues present on both days. Incomplete longer totals stay unavailable.
Uniswap V4 supplies 14.6% of all tracked daily trading. Without it, the remaining venues record $5.4B. That is the amount left after removing the leader, not an estimate of fresh money.
Selected past incidents show why the exact product matters. These are historical records, not a list of currently vulnerable pools or a complete incident history.
A compromised administrator account allowed funds to be taken from eight pools. Raydium said its concentrated-liquidity pools and staking programs were not affected by this incident. Administrative control was the problem; trading activity would not have exposed it.
Read Raydium’s incident account ↗The exploit affected Elastic’s pool accounting. That is a different product from the aggregator that finds routes across exchanges. An incident attached to one product must not become a claim about every product under its name.
Read KyberSwap’s incident account ↗The incident affected particular Composable Stable Pools in V2. Balancer’s account says V3 was not affected by this attack. A shared brand does not mean shared vulnerable code.
Read Balancer’s incident account ↗These incidents are not automatically assigned to token-search results. That requires the same network, deployed contract and affected version.
Decentralised exchanges introduce different questions: who can change the code, whether a pool can handle your trade, and what a past exploit affected. Trading volume cannot answer those questions.
Did an incident affect this deployed version, a different pool, the website or a connected bridge? The name alone is not enough.
How far would your trade move the price? Total pool value can look large while the usable liquidity is small.
Can someone upgrade the contract, pause withdrawals or alter fees? A code review does not remove those powers.
Was the fault fixed, was a new contract deployed and were users repaid? An old exploit and an unresolved exploit mean different things.
This page currently measures trading and pool readings. It does not publish an exploit-clearance or contract-safety score.
Choose the network and paste the complete contract address. The result compares returned markets by price, gross trading and available liquidity. Transactions are not wallets, and gross trading is not net money.
Research by Ananda Banerjee for Charlie Quant Lab · Updated