Monad brought almost a billion dollars onto a new chain, and collects seven thousand a day in gas

the verdict
The capital genuinely arrived: $940 million, the most since the chain has existed, and $2.02 billion traded in a month. What isn’t arriving is the till: $220 thousand of gas in thirty days, 556 times less than the same bet promised.
four measured every night, three judged every three months

What does it do, in one line?

It executes the transactions all at once instead of one after another, and assumes they don’t touch each other. On Ethereum every transaction waits for the previous one to finish, because two of them might touch the same piece of data and the order would change the result: it is a queue out of caution.

Here the opposite assumption is the starting point — most transactions have nothing to do with each other — and everything runs in parallel. Afterwards the system checks who stepped on whose toes: those few are redone, and since the data they need is already in memory, redoing them costs little. The final result is identical to the queue’s, but you get there in a fraction of the time.

all at oncethe redone ones
the transactions all start at the same instant instead of queueing; only the ones touching the same data are redone afterwards, with the data already in memory.

Transactions are executed in parallel on the assumption that they have no dependencies on each other. At confirmation the system checks which ones read or wrote the same data: only those are re-executed, and the cost is low because the data needed is already cached. The final outcome matches that of sequential execution.

The second idea is simpler still: separate the agreement on the order from the execution. While the network is agreeing on what goes into the current block, it is executing the one from three blocks back. That way the execution has the whole block’s time to work in instead of a slice, and nobody waits for anybody. All of it stays compatible with Ethereum’s code: anyone who has already written a contract brings it here without rewriting it.

01 · measured

How much adoption does it really have?

$940 million of capital, and it is the most since the chain has existed: three months ago it was $411.2 million, so 129 percent more in a quarter. For a chain that launched on 25 October 2025 it is the part of the job almost nobody manages.
$411.2mthree months agohalfway$940mtoday
the capital on the chain, from three months ago to today. source: defillama-chain-tvl, 2026-08-21.

The capital deposited on Monad went from $411.2 million three months ago to $940 million today, which is also the chain’s all-time high: 129 percent more in a quarter.

The capital is also good quality: $713.2 million of it is stablecoins, 76 percent of the total. That is a substantial difference from the chains that count their own token as capital: what is in here is money worth the same everywhere, which means it could also leave for anywhere — and that makes its staying a fact rather than a hope.

The applications with capital on them number 133, and they aren’t all clones: there are the big lending and yield protocols that sit elsewhere too, and there are things born here. It is an ecosystem, not a shop window — the real question, which is in the next section, is whether any of it produces anything for the chain.

02 · measured

Who pays to keep it running?

This is where it all falls down: $220 thousand of gas in thirty days, about $7,333 a day, which is $2.7 million a year at the current rate. The same source that told the story of the bet in 2025 did the opposite sum: ten thousand transactions a second at half a cent each makes a billion and a half dollars of fees a year.
the estimate · $1,500mtoday · $2.7m
the gas the chain collects in a year at today’s rate, against the billion-and-a-half estimate that circulated at launch. source: defillama-chain-gas30, 2026-08-21.

At the current rate the chain collects about $2.7 million of gas a year. The estimate that accompanied the launch — ten thousand transactions a second at half a cent each — worked out at about $1,500 million a year: 556 times as much.

Between the estimate and the reality there is a factor of 556. And it isn’t that the traffic is missing: on top of the chain the applications collected $6.15 million in thirty days, thirty times what the chain collected. The money is being made, but the protocols make it — the road stays free.

It is the structural defect of this generation of fast chains: passage costs almost nothing by choice, so the value moves entirely to whoever builds on top. While the token lives on expectations that is fine; the day it has to be worth what the network collects, this number is the number.

03 · measured

How much trades up there?

A good deal, and it is the healthiest number on the page: $2.02 billion traded in thirty days, which is 2.1 times all the capital sitting on the chain. Rotation like that means the money is working instead of parked.
2.1×
the chain’s capital,
traded every month
a month’s volume divided by the capital on the chain. source: defillama-chain-swaps30, 2026-08-21.

In thirty days $2.02 billion went through Monad’s DEXs, against $940 million of capital deposited: every month 2.1 times all the capital present on the chain is traded.

With $713.2 million of stablecoins on it, the depth is there for serious orders, not only for trades of a few thousand dollars. It is the difference between a market you can get into and out of and one you only get into — and it is also why the big protocols agreed to come here.

The caution to keep is that high volume on young chains is often bought: incentive programs, points, campaigns. Some of these trades are the consequence of those programs, and nobody knows how many until they end. The way to find out is to look at this number three months after the incentives stop, not before.

04 · measured

What does what is there depend on?

On the first application for 33%, which is a lot but isn’t everything: behind it there is a real tail of 132 applications, and the top five hold hundreds of millions each. That is the difference between a blockchain and an application with an address.
the first · 33%the other 132
the first application and the tail of the ones behind it. source: defillama-chain-first, 2026-08-21.

The first application holds 33% of the chain’s capital; the other 132 share the rest, with the top five worth hundreds of millions each.

The concentration that remains is the typical one: the lending and yield protocols that sit on every big chain arrive first, bring capital they know how to manage, and take the leading positions. It is capital on rent — mercenary, as the word goes — that leaves for wherever the yield is better.

So the thing to watch isn’t today’s number but what it is made of in six months. If the tail lengthens with things born here, the chain has built itself an ecosystem; if the top five stay the same five that sit everywhere, Monad is a place capital passes through, not one where it lives.

05 · judged

Who is in charge here?

Whoever puts the token up as security. This is a blockchain, not an assembly: you don’t vote on market parameters the way you do in lending protocols, you choose which version of the program to run. The power sits with the validators and with whoever delegates their token to them, and the decisions that count go through software upgrades, not through a written proposal.

The token has two declared uses and neither of them is voting: you pay gas with it and you put it up as security to keep the network running. It is a standard, honest model, and it carries no right to the revenue — and the revenue, as two sections ago showed, is small anyway.

paying gasyes
security for validatingyes
a right to the revenueno
a vote on parametersno
what the token is for, and what it isn’t.

The MON token is used to pay gas and to be put up as security by the validators for the network’s safety. It confers no right to the revenue and no vote on market parameters: the decisions go through software upgrades adopted by whoever validates.

The critical note the 2025 piece raised was about what hadn’t been published: how much of the supply was in the hands of the team and the investors, and on what unlock schedule. Until those schedules can be verified from outside, whoever holds the token carries a risk they cannot measure — and that weighs on the judgment more than any number on this page.

06 · judged

Does the team hold up?

Yes, and on this criterion the judgment is the highest on the page. The chain launched on 25 October 2025 and has not stopped running since: no public outage, no hours-long halt of the kind other fast chains have had in their first years. For a system that executes in parallel and separates consensus from execution, that wasn’t the obvious outcome.

In the meantime they have brought 133 applications with real capital up here, and nearly $940 million, in a year when persuading anyone to move away from where they already have users and liquidity is the hardest thing in the sector. Compatibility with Ethereum’s code played its part, but on its own it isn’t enough: there are dozens of compatible chains, and nearly all of them are empty.

What is left to prove isn’t technical. It is whether this capital stays when the incentives end, and whether the chain will find a way to keep some of what it generates. Those are two tests you pass with time, and the time here has only just started.

the mainnetlive since 25 october 2025, with no public outages
the ecosystem133 applications with capital on them, not only clones
the capitalat its all-time high, 76% in stablecoins
left to provethat it stays when the incentives end

The mainnet has been running since 25 October 2025 with no public outages; 133 applications hold deposited capital; the capital is at its all-time high today with $713.2 million in stablecoins. What is left to prove is whether the capital holds after the incentives end, and whether the chain can keep part of the value it generates.

07 · judged

Where does it break?

The first risk is the one the numbers spell out: a chain collecting $2.7 million a year is worth what it promises, not what it produces. If the market stops paying for promises — and it does that every cycle — the token finds itself with no floor, while the network keeps running perfectly well.

The second is mercenary capital. Nearly a billion dollars arriving in ten months arrived partly because it paid to: when the incentives end, the speed it came in at is the speed it can leave at. The chains that survive that passage are the ones where something was born in the meantime that doesn’t exist elsewhere.

put bluntly
persuading somebody to switch blockchain is like persuading them to switch supermarket when the current one is downstairs and they already know where everything is

Put bluntly: persuading a developer to switch blockchain is like persuading them to switch supermarket when the current one is downstairs and they already know where to find everything.

The third is the crowd: fast Ethereum-compatible chains come in a queue, and each one has to persuade developers to leave a place where they already have users and liquidity. Monad is better placed than most — the capital and the volume on this page prove it — but it is running the same race as twenty others, and there is only one prize.

the direct competitors, today
criterionmegaethmonadbasearbitrum
capital on it$44m$931.9m$5.00bn$1.33bn
gas, 30 days$52k$211k$1.8m$310k
traded, 30 days$37.4m$1.81bn$18.64bn$3.60bn
stablecoins on it$20.1m$711.8m$4.92bn$3.41bn
in the first app64%34%71%39%
the same measured criteria on the direct competitors, all taken today from the same source: defillama, 2026-08-20.

Comparison between monad and the direct competitors in the same category, on the measured criteria and all taken on 2026-08-20 from the same source (defillama). Capital on it: megaeth $44m, monad $931.9m, base $5.00bn, arbitrum $1.33bn. Gas over 30 days: megaeth $52k, monad $211k, base $1.8m, arbitrum $310k. Traded over 30 days: megaeth $37.4m, monad $1.81bn, base $18.64bn, arbitrum $3.60bn. Stablecoins on it: megaeth $20.1m, monad $711.8m, base $4.92bn, arbitrum $3.41bn. In the first app: megaeth 64%, monad 34%, base 71%, arbitrum 39%.

how it updates

this review updates itself, within declared limits

the 4 measured criteria are recomputed with the same procedure as on day one. the 3 judgments are not — those are revisited by hand.
rhythmcapital, fees, gas, stablecoins, swaps and the first app’s share every night · judgments every three months early triggerthe review is triggered early if the capital halves, if the gas collected in a month passes a million, or if the first application passes half the total nextby November 2026 checkedAugust 19, 2026 last recomputedSeptember 9, 2026
the words in this piece · 12
blockchain
a register of entries that sits on many machines at once, where every block carries the fingerprint of the one before it, and rewriting the past costs more than it pays.
fee
what you pay to use a protocol. it can go to whoever supplies the service, to whoever holds the token, or to both.
gas
what you pay the network to have an operation carried out: it changes with the traffic, not with the value of what you are doing.
governance
the set of rules by which decisions get made about a protocol: who proposes, who votes, who executes.
lending
borrowing onchain: you leave one coin as collateral and have another lent to you, at a rate that rises and falls with demand.
proposal
the formal proposal that gets voted on: the text, the parameters and the code to run if it passes.
stablecoin
a token built to be worth the same as a currency, usually the dollar. what changes is how it manages that: reserves at a bank, collateral onchain, hedges on derivatives.
supply
how many tokens exist. it can be the amount in circulation or the maximum possible.
token
the unit a protocol issues. it can serve to vote, to pay, to receive revenue, or to do nothing at all.
tvl
the total value deposited inside a protocol. it measures trust, not revenue.
unlock
the moment when tokens locked up until that day become sellable.
yield
what a deployed capital earns, written as a yearly percentage.
review · monad · reassessed August 19, 2026all the reviews