Unichain is no longer shrinking uniformly; it is splitting into a quiet protocol layer and a still-active network layer, and that split will decide whether the chain survives as a niche venue or fades further.
DefiLlama, which tracks deposits in decentralised finance protocols, put Unichain's locked value at $23.7 million on 12 November 2026. That figure sits 88.4% below $204.4 million a year earlier and 97.4% below the $902.9 million peak. growthepie, which measures activity from on-chain records, covers the week to 11 November, and its figures show why the deposit collapse alone misleads.
Ether prices cannot account for the deposit fall
On 13 October 2026, the locked total stood at $26.4 million, with the latest reading 10.0% lower. Ether, the network's main asset, was $2,055 on the latest date, against $2,446 on the October date, a fall of 16.0%. The token therefore dropped faster than the deposit total over that month, which points to deposits or a different mix of assets cushioning part of the decline.
Mid-August tells the opposite story. The locked total was $27.3 million then, and the latest figure was 13.0% lower. Ether was $1,884 then, and the latest price was 9.1% higher. A rising token price should have lifted the dollar total on its own, so a fall over the same stretch points to money leaving the tracked protocols rather than to price weakness.
Separating price from participation matters because the deposit figure moves for two reasons at once. A higher ether price can raise the dollar total without a new deposit, and a lower price can shrink it without a withdrawal. Those two comparisons show both effects at work, which is why the headline total needs the activity figures beside it.
Transactions fall sharply while addresses climb
Transactions averaged 302,835 a day in the latest week, down 53.3% from 648,136 thirty days earlier. The August week was busier still, at 744,347 a day, leaving the latest figure 59.3% lower. Use has therefore thinned by more than half on both comparisons, not by a small margin.
Active addresses averaged 2,801 a day, up 20.2% from 2,332 a month earlier, and up 27.9% from 2,191 in the August week. One address is not one person, so the rise records more distinct addresses taking part, not a headcount of users.
Together, those series describe lighter, more scattered use rather than a sudden departure. Transactions fell while addresses rose, which fits a network where existing participants transact less often or where new addresses arrive for small, occasional activity. Neither pattern proves that the network has found a stable base, but both cut against treating the transaction fall as proof that nobody is left.
Dollar holdings stay steady as daily payments edge up
Dollar-pegged holdings have not followed the deposit total down. growthepie measured stablecoin market capitalisation on Unichain at $28.8 million on 11 November, up 6.2% from $27.2 million on 12 October. The mid-August total was $26.8 million, leaving the latest figure 7.6% higher.
Their steadiness matters because these tokens aim to keep a steady dollar value, so their combined total is not tied to moves in ether. Dollar holdings stayed on the network even as assets in tracked protocols shrank, and the two totals measure different things: supply on the network, and deposits in its protocols.
Fees averaged $283 a day, up 15.7% from $244 over the previous month, but down 18.7% from $347 in the August week. People were still paying to transact, and paying more than a month earlier, even though they transacted far less often.
The test ahead is whether addresses keep coming
Both sources have clear edges. DefiLlama's locked value covers only the protocols it tracks, and that figure shifts with token prices and with money moving in or out. growthepie counts activity and stablecoin supply on the network itself. Smoothing over seven days hides short bursts, and an address is not a person, so neither series can say how many people remain or why they transact less.
That shape matters for anyone deciding whether to build or transact here. A network with more addresses and fewer transactions can still support occasional use, but it offers thinner liquidity for large deposits and less revenue for protocols that depend on locked assets. The deposit figure, in that light, is less a verdict on the chain than a measure of one layer of it.
The useful test for people who use or follow Unichain is not whether locked value returns to its peak, but whether the rise in addresses holds while transactions stop falling. If transactions stabilise beside steady dollar holdings, the smaller network can be read as a core that stayed; if addresses turn down as well, the thinning becomes an exit. Future figures will show which account the data supports.
Not investment advice.
Source: DefiLlama, growthepie, Unichain fundamentals · Date: