Nexus Market: the multisig pitch that ended in a 2026 exit scam
Updated
Nexus Market is a Tor market that appeared in late 2023, takes Monero by default with Bitcoin as a second option, and advertises 2-of-3 multisig escrow and daily mirror rotation as its main features. It was mid-sized and known mainly through its own promotion, copied across a network of clearnet sites. Between about 20 and 30 September 2026 withdrawals were blocked and users could not get their balances out: an exit scam. Our read: none of its security claims was ever tested in public, and the multisig that was its main selling point did not stop the loss for anyone who left money on the market. Anyone searching for a Nexus market link now is most likely to find the promotional network, which is still online.
September 2026: withdrawals blocked
In the last ten days of September 2026, around the same time as DarkMatter, Nexus stopped letting users withdraw. Balances stayed locked and were not returned. No police agency has announced a seizure, so by the checks on the darknet news page this is an exit scam, not a takedown. Treat any balance left on Nexus as lost, and treat any “Nexus is back” or “withdrawals reopen” post with suspicion: they are the usual bait of phishing copies. The rest of this page describes the market as it was advertised before the closure.
Not the other Nexus
Nexus is one of the most reused product names in tech, so a bare search mixes several unrelated things. Sonatype Nexus Repository is the artifact manager that stores Maven, npm, and Docker packages for software builds; if you came from a DevOps query, that is the one you want. Cisco Nexus is a line of data-centre switches. Google sold Nexus phones until 2016. None of them has anything to do with the Tor market, and the market has no connection to them beyond the word.
Nexus market link sites: one story, copied dozens of times
Search for Nexus Market and you get page after page telling the same story in nearly the same words: launched November 2023 after a closed beta of about 200 buyers, multisig “as a default rather than an opt-in”, Monero by default, mirrors rotating every 24 hours, and a signing key that “has not rotated since the platform opened”. The sites take different shapes. One presents itself as an onion encyclopedia, one as a “verified gateway”, one as a review site with a 4.7-star score from more than a thousand reviews, none of which you can read. Two of them carry the same update timestamp to the minute.
That is a marketing campaign. Promoting a market this way is normal; most markets pay or encourage link sites. What is unusual about Nexus is that nothing else exists alongside it. Bigger markets like Torzon also turn up in security research, enforcement news, and years of forum threads. Nexus turns up almost only in its own copy, and the promotional sites do not even agree with one another on which addresses are current.
Scale figures follow the same pattern. Its promotion claims about 4,600 vendors and 28,750 listings; the working catalogue is nearer 4,000 listings.
The short key ID is not a fingerprint
Nexus promotion identifies the market’s PGP key by an eight-character hex ID, written 0x7F2A0A9D. That looks like a security feature and is close to useless as one. An eight-character ID is the last 32 bits of the key’s fingerprint, and generating a different key with the same 32 bits takes seconds to minutes on an ordinary computer. In 2014 researchers did it for tens of thousands of keys in the public web of trust to prove the point.
A real PGP fingerprint is 40 hex characters. Checking a market’s signed mirror list means importing its public key from a source you already trust, comparing all 40 characters, and then verifying the signature on the message. A short ID proves nothing about who signed a list. If a phisher made their own key with the same short ID and signed their own mirror list, a reader who checks only the eight characters would accept it.
What real 2-of-3 multisig would look like
Multisig is the one Nexus claim that would matter if true. In a 2-of-3 setup the buyer, the vendor, and the market each hold a key, and any two of them can release the payment. The market alone cannot move the coins, so an operator who decides to exit cannot simply take the escrow balance. It is the strongest custody design a market can offer, and the reason it is rare is that it is fiddly for users.
That fiddliness is also how you tell real multisig from a banner. A genuine 2-of-3 order needs something from the buyer: a public key, a refund address, or participation in setting up the multisig wallet, which on Monero takes several rounds of exchanging data. If a market’s checkout never asks the buyer for anything, the coins are sitting in a wallet the market controls, whatever the front page says. Nexus never showed how its multisig worked, and the blocked withdrawals in September 2026 fit market-held escrow rather than a design where the operator cannot move the coins. How the two models differ in an exit is on the darknet economy page.
Where Nexus stood against the other markets
Nexus was not a newcomer like Black Ops, it was not under the scale pressure of Torzon, and it promised the custody protection that DarkMatter documented. Monero by default put it in the same camp as DrugHub and DarkMatter, the privacy-first end of the market. DarkMatter exited in the same weeks; DrugHub is still trading.
No enforcement action means no seizure, no arrested admin, and no public leak tied to Nexus. It also means no one outside the market ever checked whether the multisig, the key handling, or the mirror rotation worked the way the copy said, and the exit shows the claims did not protect users. The risk that remains is the promotional network around it: a search result that copies the story perfectly and serves a different address, now with a closed market behind it. The markets overview explains how those phishing copies work.
What the record on Nexus can and cannot show
No agency has published anything about Nexus, so this page rests on the service’s own promotion and on the sites that copy it, and we could verify none of their claims. The exit is an assessment, not an announcement: withdrawals stopped in late September 2026 and no police agency claimed the closure. We have not used the service. Our editorial standards describe how we word that difference.