Iran’s Crypto Exchanges Claim 45 Million Accounts. One Holds a Current Certificate.

Eleven Iranian crypto exchanges publish an account or user number on their own websites. Added together the eleven come to about 45 million accounts. Tehran Index holds 22 crypto exchange records and checked every one against the state eNamad database: four carry the state's general e-commerce trust seal, which is not an exchange licence, three of those four have expired, and one is current. Seven of the 22 say nothing about their own standing on the pages we read, including the two most installed apps in the sector. One exchange does have audited accounts in public, and only because a listed venture fund owns 2.6 percent of it. This measures what these companies disclose. It is not a finding about anyone's compliance with a rule this desk has not read.

Eleven Iranian crypto exchanges publish an account or user number on their own websites. Add the eleven together and you get about 45 million accounts. Tehran Index holds 22 crypto exchange records, and across all 22 there is one current certificate in the state eNamad database: the electronic trust seal issued to Iranian e-commerce sites generally, not an exchange licence, and it belongs to none of the four largest.

That is not an accusation and it is important to say so immediately. We looked for a published list of authorised Iranian crypto exchanges, at every regulator and trade body we could identify, and found none. Not finding a list is not the same as establishing that no rule exists. Tehran Index has not read any regulation governing this sector, does not assert what one would require, and does not say that any company named here has fallen short of anything. What follows is a measurement of what these companies disclose about themselves and what an outside reader can check.

The one certificate

The one state database this desk checked all 22 against is eNamad, the electronic trust seal issued to Iranian e-commerce sites generally. It is not an exchange licence. It is included here for one reason: it is a state issued record in which a crypto exchange and a verifiable credential can appear together, and anyone can look it up.

Four of the 22 carry one for the domain on their record. Three of the four have expired. OmpFinex is the only company in the sector holding a current one.

One of the four is a different shape from the rest. MyWallet’s certificate is held by a private individual rather than by a company. We are not suggesting anything is wrong with that, because we have not read the seal’s eligibility rules and do not know what they permit. It is worth recording only because every other holder in this set is a registered company, and because it means the certificate does not identify an operator.

And one company states something the state database does not carry. Bidarz prints on its own site that it holds an electronic trust seal. A query against enamad.ir for its domain returns no certificate. The site prints no certificate reference, so there is nothing to look up. We are not calling that claim false: a seal can sit on a different domain, and a database read is weaker than the thing it probes. We are saying that a reader who tries to verify it cannot.

What the companies say instead

Nine of the 22 state something about their own standing. What they name is a mixed set, and not one of them is a record specific to running an exchange. Tabdeal, Exir, Pool-e No and OK Exchange each associate themselves with the knowledge based company register, which is a technology classification. MyWallet and Sarmayex list memberships of trade bodies. OK Exchange’s home page calls it a site that holds a permit and names no issuer; its licences page cites the knowledge based designation. We have not read the criteria behind any of these, and we are not ranking them against each other. The point is narrower: none of them is a record in which a reader could confirm that this particular company is authorised to do this particular thing.

Ramzinex goes furthest. Its meta description, which is the sentence it offers search engines to display, states that it is an authorised Iranian exchange approved by the cyber police and the central bank. Neither of those bodies publishes a register this desk could find in which that approval would appear. The claim is recorded here as the company’s own words, in the company’s own place, and it is not adjudicated. OmpFinex is more careful with the same two names, saying only that users must supply identity information under those bodies’ instructions, which is a description of its own onboarding and not a claim about itself.

Seven state nothing about their standing on the pages we read, and the group includes the two largest companies by install bucket: Nobitex, which publishes the most precise account figure of the 22 and says nothing about its own authorisation on those pages, and Bitpin. There is a case for reading that silence as the more careful posture, since a company that claims nothing is not asking a reader to believe anything it cannot check. There is also a case for reading it as a gap. This piece does not choose between them; it records that the sector’s two most installed apps leave the question unanswered.

The remaining six are: Excoino, whose expired certificate is the only credential we found; Bit24, which claims nothing on its site; Aban Tether, whose site could not be loaded; Arnika, a holding company whose record domain does not resolve and may be mistyped on our record; and Nobidex and Tidex, which are products of Nobitex and Tabdeal rather than companies in their own right.

Who these companies actually are

Seventeen of the 22 are anchored to a registered legal entity on our register and five are not. That is 77.3 percent, against 86.4 percent across all 369 public records, where 319 are anchored.

The obvious headline does not survive the data, so here is the number that kills it: crypto is not the worst anchored sector on this register. Among the sectors where we hold at least eight public records, gaming is 5 of 8, health 8 of 12, artificial intelligence 11 of 15 and accelerators 6 of 8, every one of them below crypto. This sector sits below the registry average on legal identity and it is not an outlier, and anyone reporting it as uniquely opaque on that measure would be reporting it wrongly.

The individual cases are still striking. Bit Barg publishes a claim of more than two million accounts and names no company on any of the three pages a reader would look at. Its terms of service name none, its about page names none, and its footer prints a copyright line and nothing else. A research pass on 21 September looked for a registered entity in the Official Gazette and found no candidate to offer.

Bit24 is the more instructive case, because a candidate exists and we are not publishing it as an answer. A Gazette entity’s name contains the developer string one of the app stores prints, and three third party profiles assert that it owns the Bit24 brand. None of the four primary links this desk requires holds: bit24.cash names no company on its about page, its rules page or its footer; the domain carries no eNamad; and the Gazette notices name the company without naming the brand or the domain. A store developer string is a lead. Resemblance is not evidence, and this register has been stripped of rows that rested on it before.

Why nobody can check the 45 million

The eleven account claims are not one quantity. Nobitex publishes a count to the person, 11,490,676 at the end of 2025. Tabdeal publishes more than eight million, Bidarz more than two million and Pool-e No more than one million, and all three label theirs active users. Wallex publishes more than five million users and, separately, more than one million active. Ramzinex publishes five million and four million active. Ten of the eleven figures are floors written with a plus sign rather than counts. Adding them produces about 45 million accounts, and that number is not a population, because one person holds several accounts and the companies are not measuring the same thing.

For most of the 22 the only outside figure about scale is the app store install bucket, and three have no app at all. It cannot do this job. Our own earlier work established what that field is: a stale bucket that barely moves, where the rater count is the only value that behaves like a series, and where the two Iranian stores disagree about every app they share. An install is not an account and an account is not a person. We do not subtract one from the other and nor should anyone else.

What can be said without netting them is narrower and still worth saying. For Bidarz the published account claim is more than two hundred times its install bucket, and for Bit Barg more than twenty times. Bit Barg’s site also prints a users figure of 1,000,000; on that figure the ratio is about ten. Those are the two widest gaps in the sector; the next is Pool-e No, whose claim is about thirteen times its install bucket. Neither Bidarz nor Bit Barg publishes anything that reconciles the two figures, and both could.

The one exchange whose accounts you can read

There is a single exception in the sector, and the way it arose is the whole argument of this piece in one company. Wallex has audited revenue, audited profit and an audited balance sheet in public for two consecutive years. None of it comes from Wallex. The Sarv Roshd Paydar Yekom venture fund owns 2.60 percent of the company, the fund is a listed issuer, and its auditor prints each portfolio company’s own accounts in a note. That note is the only public source for these figures this desk has found.

Revenue was 327.1 billion toman in the year to 20 March 2025 and 484.0 billion in the year to 20 March 2026, a rise of 48.0 percent in rial terms with no inflation restatement applied. Net profit went from 25.4 billion toman to 30.0 billion, up 18.1 percent. Equity went from 299.2 billion to 809 billion. The fund carries its 2.60 percent at a cost of 25.0 billion toman (about $0.14M at ~178,000 toman/$), and that is a cost and not a price: Tehran Index does not gross a minority cost up into a valuation of a private Iranian company, and nobody else should either.

Convert each year at its own period rate and the growth goes the other way. Revenue of 327.1 billion toman is about $4.7M at the audited 2024-25 average of roughly 70,000 toman to the dollar. Revenue of 484.0 billion toman is about $2.7M at the 2025-26 filings rate of roughly 178,000. So a business that grew 48.0 percent in rial shrank about 42 percent in dollars, and net profit of about $0.36M became about $0.17M. The two rates come from different bases and the comparison should be read as an order of magnitude rather than a precise series, but the direction is not in doubt and it is the same direction this desk found on the Tehran market. Anyone quoting the 48 percent without the currency is quoting half the sentence.

Now put that beside the company’s own marketing. Wallex publishes more than five million users in its hero and more than one million active users in its body copy, a total and an active figure with no period on either, and the Tehran Index record has carried them since 18 August 2026 as the company’s own words and not as an operating figure. So the one exchange on this register with audited accounts anyone can read is also an exchange whose own user claims this desk reports only as the company’s words and does not carry as an operating figure. The numbers we can stand behind came from a third party with a filing obligation. The ones we can only quote came from the company.

What we do not know

Whether a licensing regime exists for this sector. We searched for a regulator, ministry or trade body publishing a list of authorised Iranian crypto exchanges and found none. That is the absence of a list we could find. It is not evidence that no rule exists, and this piece makes no claim about what any rule requires or about whether any company complies with one.

Two of the 22 were not read. abantether.com returned a redirect loop and ernyka.com does not resolve. One is a redirect loop and the other a record domain that may be our own typo; neither is a statement about the company behind them. A failing probe is not a finding.

The certificate check was run against the domain each record holds. A company holding a certificate on a different domain would not appear. The Ramzinex record has since changed from ramzinex.com to ramzinex.ir, the domain Cafe Bazaar prints, and that domain has not yet been checked.

Two of our own reads disagree and we are publishing that rather than picking one quietly. The research file and the App Index capture read the same Cafe Bazaar listings on the same day and agree exactly on 17 of 19. They disagree on Nobitex, 11 million by hand against 5.4 million in the capture, and on Bitpin, 6 million against 2.6 million. The table above uses the capture, because that is the gated one. Neither figure should be treated as settled until those two listings are read again.

The Gazette work behind the identity section was done through a mirror of the official notices rather than the notices themselves, because the registry site returned a redirect loop. The one identity this pass resolved, Hitobit, has not yet been loaded to the register, so the count of five unanchored records should become four once it is.

And for 21 of the 22 we do not know what the business earns. We found no accounts published by any of the 22 themselves. The one exception is covered above and it is an accident of somebody else’s shareholding, not a disclosure any of these companies chose to make.

New in this piece, against what came before it: Iran’s Crypto Adoption, in Global Context used one company’s claimed reach to place Iran on an international adoption chart. This is what that number rests on, and the answer is the company itself and nothing else. The Name Is Not the Company showed how far a brand sits from its registered operator across the registry. This is a sector where that method reaches its limit. And Most of Iran’s Venture-Backed Companies Make a Profit counted what venture funds disclose about their portfolios; the Wallex accounts above are that mechanism arriving in a sector that has no other route to a public number. If you are handed a user number for an Iranian crypto exchange, there are three things to check before using it: whether the company names the entity behind the brand, whether the figure is accounts or actives, and whether anybody other than the company has ever published it. In this sector we found no case where the figure originated with anyone but the company.

Cite as: Tehran Index, “Iran’s Crypto Exchanges Claim 45 Million Accounts. One Holds a Current Certificate.”, 25 September 2026. Built from the 22 companies’ own websites, read 21 and 22 September 2026, the state eNamad database, read 21 September 2026, and the Tehran Index company registry and App Index queried live on 25 September 2026. tehranindex.com