Is Algorand dead? What the chain shows in 2026
Algorand carries more than 20 million transactions per month and counts its active wallets by the hundred thousand. We sorted both by where they come from, all the way back to the launch in 2019.
- 32%
- 374,234
- 170,621
- 5,082
- 88%
- 150x
01 · The jumpThe month Algorand's wallet count rose 30%
In August, the number of active wallets on Algorand jumped 30%. The Algorand Foundation, the body that looks after the chain, reported it in its monthly review. It said participation had "strengthened materially".
We sorted the same wallets by what they were doing on-chain, and the whole rise came from two places.
Of every five wallets added, three came from a payments app in Afghanistan. The other two came from a scam. It sprays dust, payments too small to matter, from fake wallets, and both the fakes and their targets count as active. All other wallets, taken together, fell 12%.
Algorand is a layer 1 blockchain that went live in June 2019. Silvio Micali, a cryptographer at MIT, designed it to be quick and cheap, so a new block lands every 3 seconds or so and sending a transaction costs about a hundredth of a cent. It has its own coin, ALGO, which peaked above $3 in the month of the launch and now trades at about 13 cents.
So is Algorand dead? Bitquery indexes every Algorand block, so we read them all, from the first one to the end of this September. The chain is alive. Blocks land on time, and repeat use of an Afghan payments app has more than quadrupled in a year. But most of what gets counted as activity comes from a few programs. The biggest is a bot that copies chess games, and behind it come an app opening accounts in bulk and a scam. Take out the app and the scam and about 5,000 wallets use Algorand on 5 or more days in a month, half as many as 3 years ago.
02 · Why people askA chain with busy years behind it
Two years ago, in its busiest year, Algorand carried about 49 million transactions per month. This year it carries about 23 million. Anyone asking whether the chain is dying is usually looking at a chart like this one.
Two honest counts of the same chain can be far apart. When we say transactions, we mean the ones somebody sent. The Foundation's totals also include the extra steps an app runs inside a transaction, which is why its August figure, 32.9 million, is larger than ours, about 20 million. Its wallet count is wider too. It counts any wallet that sent or received something, while we count the ones that sent, since anyone can be sent a token they never asked for.
03 · The chess logA third of the traffic is copies of chess games
The busiest wallet on Algorand does one thing. About 3 times per second, it sends itself a token that is worth nothing, with a short note attached. The note holds the result of a chess game.
Last month the wallet sent 7.7 million of these, about a third of everything sent on Algorand. By the Foundation's wider count the log is still a fifth.
The log is nothing new. Four years ago it averaged about 6 million results per month, and this year it averages about 7 million. Its share has climbed mostly because the traffic around it has shrunk.
The games are real. They were played on Lichess, a free chess site. We pulled 49 games from Lichess's public records and found 45 of them on Algorand, with the same ratings, the same points won and lost, and the same starting second. The players' names are scrambled, though in a way that always turns the same name into the same string.
Nothing we found shows that Lichess has anything to do with this. It publishes its games for anyone to download, so copying them takes nobody's permission.
So who is copying them? The trail starts with World Chess, the company that runs the official online platform of FIDE, chess's world governing body. In 2020 the platform was called FIDE Online Arena, and that June Algorand and World Chess announced that its ratings would be recorded on Algorand. One wallet did that for the rest of the year, logging tens of thousands of results under tokens named FIDE Online Arena. The next February the same wallet switched to tokens whose names begin LC and started posting Lichess games, hundreds of thousands per day. A second wallet took over in late 2023 and is still going. We cannot say who runs it today.
Since late 2021 the log has never been less than six weeks behind, because Lichess hosts far more games than one wallet can post.
On one day we checked in late September, every game the wallet posted had started inside the same 52 minutes of one day in early July. A day of posting covers an hour or two of play, so the log falls further behind every day. In March it was still posting games from the June before. Every so often it skips ahead and starts again.
Between them the two wallets have sent 340 million transactions, each one costing a thousandth of an ALGO. Last month the log paid 18% of all the fees Algorand collected.
The smaller bars on that chart include a scam and a bot that warns people about the scam. Another bot calls one app once in every block. There is also a newer kind of traffic, x402, a way for software to pay software a fraction of a cent at a time. It made hundreds of thousands of payments last month, worth about $41,000 in total.
04 · The walletsHalf of September's active wallets were one app's new accounts
On-chain, a wallet shows up as an address, a long string that works like an account number. More than half a million of them sent at least one transaction in September, and two sources account for most of them.
The first is HesabPay, a payments app in Afghanistan. Its token, the Hesab Afghani, stands in for the local currency, and aid agencies have used it to deliver cash.
HesabPay made up half of the month's active wallets, and most of those were brand new. Across August and September the app set them up by the hundred thousand and gave each one about a fifth of an ALGO to cover fees. When we looked again in early October, more than nine in ten had still not received any money. We do not know what they are for.
The bars are new wallets and the line is those that paid someone. The app has opened them at this rate once before, in 2022, when it first moved onto Algorand.
| HesabPay on Algorand, September 2026 | Wallets |
|---|---|
| New wallets set up since 1 August | 374,234 |
| Of those, had received money by 2 October 2026 | 22,515 |
| Paid someone during the month | 44,085 |
| Active on 5 or more days | 17,493 |
| The same count a year earlier | 3,927 |
The app's steady use is smaller, and it is real. Tens of thousands of its wallets paid someone last month, moving about $65 million once the very largest transfers are left out. The number of wallets that use it on 5 or more days has more than quadrupled in a year.
The second source is a scam called address poisoning, which we have measured on Ethereum. One wallet pays for thousands of fake addresses. Each fake copies the first 4 characters of an address its target has really paid. It then sends the target a millionth of an ALGO, so the fake shows up in the target's history, ready to be copied by mistake.
The scam comes in waves, and September's was the biggest yet. Its fakes made up a third of all the addresses active on the chain.
| The look-alike scam, September 2026 | Count |
|---|---|
| Look-alikes active during the month | 170,621 |
| Dust payments sent that month | 752,729 |
| Look-alikes funded since September 2025 | about 504,000 |
| ALGO collected from other wallets in those 13 months | 3,827 |
| ALGO paid in fees over the same months | 11,591 |
| Warnings sent by the WARN666 wallet | 615,993 |
The scam hardly works. Over 13 months the fakes took in a few thousand ALGO from 20 wallets, and they paid about 3 times that in fees.
Someone is fighting back. A wallet whose address begins WARN666 and ends SCAM sends targets a short message naming the fake, and it did so hundreds of thousands of times last month.
Neither source was this large earlier in the year. Leave both out and the number of active wallets fell 61% between January and September.
05 · The regularsWho comes back
Sending one transaction is enough to count as active. A better test is who comes back, so we counted the wallets active on 5 or more days in a month and called them regulars.
Two thirds of September's regulars were HesabPay wallets. Some of the rest were the scam's fakes, reused day after day. What remains is everyone else, which means every trader, app and bot on the chain, and it comes to about 5,000.
| Active on 5 or more days | Wallets |
|---|---|
| All regulars, September 2026 | 26,005 |
| HesabPay wallets | 17,493 |
| The scam's look-alikes | 3,430 |
| Everyone else | 5,082 |
| Everyone else, September 2023 | 9,703 |
| Everyone else, average month of 2023 | about 17,900 |
| Everyone else, average month of 2026 | about 7,000 |
| Everyone else, lowest month (August 2026) | 4,816 |
Three years ago there were about twice as many of these other wallets, and across the whole of 2023 the count ran at more than three times this September's level. This year the count has roughly halved, and August was the lowest month we measured.
The chart shows a bulge from late 2024 to mid-2025. In the month we tested, about a third of the bulge was wallets made to hold airline tickets. We could not identify the rest.
A few senders make most of the traffic too. The 10 busiest made more than half of the month's transactions, and the top 100 made about four in five.
06 · The boomsEvery boom was one app
Algorand's wallet count has soared several times. Each time the cause was the way one app was built.
The first was Props, a loyalty-points scheme used by video chat apps such as YouNow, Camfrog and Paltalk. It moved to Algorand in 2020, and millions of wallets appeared to hold the points. The active count went from a few thousand to almost 5 million in two months. Props shut its rewards programme at the end of 2021. In its last month 11 million wallets closed out their points and went quiet, and three months later the active count was down 94%.
The second was PlanetWatch, which paid people in PLANET tokens for running air-quality sensors. For two years PLANET changed hands more often than ALGO itself. By 2024 PLANET's traffic was down 90%.
The third showed up in transactions. In January 2024 a token called Orange could be mined by sending transactions to its app, the more the better. Algorand's busiest month followed, with 116 million of them. On the day we sampled, more than four in five of the transactions that ran an app went to Orange's app.
The fourth was TravelX, which minted airline tickets as NFTs and created wallets to hold them. At its peak it sent 3.5 million tickets in a month to 1.3 million wallets.
In June 2025 the Foundation said TravelX was leaving, and hundreds of thousands of wallets closed that same day. By July the active count had fallen by 88%.
The fifth app never touched our count at all. Alpha Arcade is a prediction market, a site where people bet on how events will turn out. From the one account we traced, it appears to open a fresh escrow account for each order, and there were as many as half a million per month. No person controls those accounts, and almost none ever sent a transaction of its own. They still appear in any count of wallets that received money. The data site DefiLlama puts the market's all-time trading at about $24.5 million.
| App | What got its own wallet | The count it pushed up | Afterwards |
|---|---|---|---|
| Props | Holders of loyalty points from chat apps | 12.2 million wallets active on the chain, December 2021 | 739,000 by March 2022 |
| PlanetWatch | Owners of air sensors, paid in PLANET | 187 million PLANET transfers in 2022 | 19 million in 2024 |
| Orange | Nobody: it was mined by sending transactions | 116 million transactions, January 2024 | 45 million the next month |
| TravelX | Wallets made to hold airline tickets | 2.6 million wallets active on the chain, March 2025 | 286,755 in July 2025 |
| Alpha Arcade | Each order, as far as we traced | 524,336 escrow accounts, November 2025 | 37 in March 2026 |
| HesabPay | Each HesabPay account | 262,273 HesabPay wallets active, September 2026 | Still running |
The pattern repeats. An app gives every user, ticket or order its own address and the count climbs, and when the app changes or leaves, the count falls back. Today's version is HesabPay's new wallets and the scam's fakes.
07 · The billThe people who run Algorand are paid by its Foundation
Thin traffic leaves one more question, which is who pays to keep the chain going. Algorand is kept running by validators, the computers that add new blocks. Since January 2025 a validator has been able to earn a staking reward for each block it adds, which started at 10 ALGO and shrinks a little over time.
Users pay fees too, and half of each block's fees goes to the validator. Next to the reward they barely register. Last month validators received about 150 times what users paid.
| Who pays for the chain | September 2026 |
|---|---|
| Staking rewards paid out | 6.49 million ALGO, about $663,000 |
| Fees paid by users | 42,758 ALGO, about $4,400 |
| Reward per block | 8.26 ALGO |
| Times the Foundation has refilled the pot since January 2025 | 12 |
| ALGO it has put in | 144.6 million |
| Accounts that added at least one block | 1,489 |
The Foundation covers the gap. The rewards are paid from one pot, and the Foundation has refilled that pot 12 times since the rewards began, from wallets it lists as its own.
Each refill brings the pot back to about 20 million ALGO, apart from one large top-up last December that lasted seven months. The Foundation first promised the rewards for about two years. In July it extended that to January 2028 and said it would change how fees are set.
08 · So, is it dead?Alive, with a thin crowd
No. A dead chain stops making blocks, and Algorand made one every 3 seconds or so all month. It has a payments app with real users in Afghanistan, and that app's returning wallets have more than quadrupled in a year.
What it lacks is a crowd. A third of its transactions are a chess log that runs months behind. Half of the wallets active last month were new accounts waiting for money, and a third were fakes. Outside those, about 5,000 wallets came back on 5 days or more, half as many as 3 years ago.
If you want to know whether a chain is being used, the transaction count will not tell you, and nor will the wallet count. Ask who comes back.
09 · How we measured itMethod
We read every Algorand block from the first, in June 2019, to the end of September 2026, and looked again in early October to see which new wallets had been paid. We counted the transactions that people and programs sent, and the wallets that sent them. A regular is a wallet that was active on 5 or more days in a month. A look-alike is an address paid for by the scam's wallet. Dollar figures use the month's average ALGO price.
The transactions, transfers and blocks docs show the fields we used. We asked the same question of Dogecoin and Cardano, and found a similar gap between traffic and people on Optimism. Polygon turned out to be mostly one app too.
10 · The recordAddresses behind this story
FAQ
Is Algorand dead?
No. Algorand added a block about every 2.8 seconds in September 2026, and a payments app built on it has more than four times the regular wallets it had a year ago. But use is thin. About a third of the transactions sent that month were copies of chess games posted by one wallet, and only about 5,000 wallets outside one app and one scam were active on 5 or more days.
How many people use Algorand?
Nobody can count people, only wallets. In September 2026, 527,746 wallets sent at least one transaction. Half belonged to one Afghan payments app and a third were fake addresses run by one scam. 26,005 wallets were active on 5 or more days.
Why are chess games recorded on Algorand?
In 2020 Algorand and World Chess announced that FIDE Online Arena ratings would be recorded on the chain. Since February 2021 the same log has posted the results of games played on Lichess, a different chess site. Lichess publishes its games openly, and nothing shows it is involved.
What is HesabPay?
HesabPay is a payments app in Afghanistan that runs on Algorand. Its token, the Hesab Afghani, stands in for the Afghan currency. In September 2026, 44,085 of its wallets paid someone, and it set up 374,234 new wallets across August and September.
How are Algorand validators paid?
Since January 2025 the validator that adds each block earns a staking reward, now 8.26 ALGO, plus half of that block's fees. Fees are tiny next to the reward, so the Algorand Foundation tops up the pot. In July 2026 it extended the rewards to January 2028.
What happened to TravelX on Algorand?
TravelX minted airline tickets as NFTs on Algorand and created wallets to hold them. The Algorand Foundation said in June 2025 that TravelX was leaving the chain. Active wallets fell by 88% the next month.
A wallet is not a person. One person can hold many addresses, and an app can create one for every user, ticket or order. That is the point of this article, and it also applies to our own count of regulars, which includes bots.
Our transaction counts cover transactions that somebody sent. The Algorand Foundation's published totals also include the steps apps run inside a transaction and the payouts to validators, so its numbers are larger. Counted its way, our August total matches its figure.
Our count of active wallets by the Foundation's definition, sent or received anything, runs 6% to 11% above the Foundation's own. We quote the Foundation's figures where we cite its report and use our own for the split.
A look-alike is any address paid for by the scam's funding wallet. Other scams of the same kind, if there are any, are not counted. The match with addresses their targets had paid was tested on a sample: 89% copied the first 4 characters.
What the scam collected counts payments our data records. ALGO swept out when an account is closed is not recorded, so the true take could be higher.
HesabPay's payment value leaves out transfers of 1 million afghani or more. The largest of those pass back and forth between the wallet that issues the token and one other. Dollar values use an exchange rate of 64.77 afghani per dollar.
Who runs the chess log today is not established. The link to World Chess rests on the token names the first wallet used in 2020 and on the 2020 announcement. Lichess is named only as the source of the games.
Two things on the charts are unexplained: about two thirds of the bulge in regulars from late 2024 to mid-2025, and a fall in other active wallets between August and September 2026.
Fees are worked out from staking rewards, since half of every block's fees is paid out with them. Adding up every fee recorded for September gives the same total to within a tenth of one percent. The balance of the rewards pot is worked back from its balance on 2 October.
This article is provided for informational and educational purposes only and reflects analysis of publicly available on-chain data as of the dates indicated. It does not constitute legal, financial, compliance, or investment advice, and nothing in it is a recommendation to buy, sell, or hold any token or asset.
The findings describe transactions and balances observed in Algorand blockchain data over the period stated, together with the rules we used to classify transactions, wallets, look-alike addresses and fees. Attributions come from token names, public statements, Bitquery's address labels and on-chain behaviour; they may be incomplete or incorrect and may be revised as more data becomes available. Algorand addresses are pseudonymous, and an unlabelled address is described by its behaviour and attributed to nobody.
References to Lichess, World Chess, FIDE, HesabPay, TravelX, Alpha Arcade, PlanetWatch, Props, YouNow, Camfrog, Paltalk, the Algorand Foundation or any other named party describe on-chain records and public statements, and are not statements about any entity's conduct, solvency or intentions. Lichess is named only as the public source of the chess games; nothing here asserts that Lichess takes part in recording them. The look-alike addresses and the wallet that funds them are not attributed to any person or company. Nothing here asserts that any named party acted unlawfully.
Nothing herein should be relied upon as a definitive determination of fact. Readers should conduct their own independent verification before taking any action. The authors and publisher accept no liability for any loss or damage arising from reliance on this material. All trademarks and company names are the property of their respective owners.
Reported by Gaurav Agarwal for Bitquery Research, with AI tools; every figure was checked against the raw data.
Count Algorand's wallets yourself
The figures above come from Bitquery's Algorand data, which holds every block, transaction and token transfer since June 2019. The same data is open through our API, so you can check any address named here or run your own counts.