A closed crypto card and where to go next

Updated October 2026
Licensed
usAvailable in US
Fast payouts
18+ Only

I went through the Moonwell card’s terms and the public notices around it, and what I found is short and blunt: the card is no longer being issued, and the wider programme behind it has been wound down. If you are reading this because you are trying to decide whether to apply, the decision has already been made for you, and it is “no.” If you are reading because you once held one and want to understand what happened to your balance, your rewards and your access, the rest of this page walks through it.

A payment card next to a phone with a crypto wallet app open
The Moonwell card programme no longer issues new cards

Conditions are current as of October 28, 2026.

The conditions described here are current as of October 2026.

Verdict

The Moonwell card closes at 2.3 out of 10, and the reason is not that it was a bad product in isolation, but that the product no longer exists in any practical sense. Card loading was stopped on 8 July 2026, the last day to use a card was 8 August 2026, and the app behind it went offline on 6 September 2026. There is no path to a new card, and even the old cards are now cancelled.

Who it suits, today, is almost nobody. The audience this page is genuinely for is the former cardholder who wants to confirm closure dates and the alternatives section, and that is the lens I have written it through.

How the score breaks down

Criterion Mark Why
Who can get it 0 / 10 Card loading stopped and new issuance ended on 8 July 2026; last day of use was 8 August 2026.
Cost 7 / 10 Unofficial figures cite a 0.5% USDC load fee and a 1.5% foreign exchange fee, but with the programme closed these are academic.
Top-up options 2 / 10 Card loading was stopped on 8 July 2026. Historically the card supported on-chain USDC and a “borrow-and-load” against Moonwell lending positions, both unofficial.
Reliability and complaints 1 / 10 The card closed together with its infrastructure provider, Cypher, which was acquired by Nium; free withdrawals ran until 6 September 2026. I did not find complaints about the card itself.
Everyday use 0 / 10 All cards were cancelled on 8 August 2026 and the app was switched off on 6 September 2026.

Two things stand out. First, the everyday-use and access scores sit at zero because the product is gone, not because it failed in a dramatic way. Second, the “cost” score of seven is a historical figure from third-party reporting, and I treat it as unofficial. Once a programme is closed, a competitive cost table is not the deciding factor, and I have not ranked alternatives against it on price.

What happened to the programme

The card did not drift away quietly. There is a dated notice from the infrastructure provider behind it, Cypher, and a brief note from Moonwell itself. Together they explain the timeline and, importantly, who is responsible for what.

On 8 July 2026 Cypher put out a notice saying it had been acquired by Nium, a global payments infrastructure company. From that day, card loading stopped and no new cards were issued. Cardholders had one month to keep using their existing cards: the last day of card use was 8 August 2026 at 00:00 UTC, after which all cards were cancelled. The final rewards cycle covered 2 to 16 July 2026, and any rewards earned before 16 July 2026 would be honoured, provided they were claimed before the end date.

Then came the longer tail. The Cypher app, the on-chain dApp, and free withdrawals all closed on 6 September 2026. From that point, withdrawing any remaining card balance was no longer free, and the app was simply offline. Cypher support, reachable at [email protected], stayed open through 6 September 2026 and then ran a 60-day transitional period for stragglers. After that, the line went dead.

Moonwell’s own message, as reported by Crypto Briefing, is shorter: the team is exploring new card providers and there is no timeline. There is no successor card from Moonwell today, and the moonwell.fi site no longer carries card content at all.

That is the bones of it. If you want the practical consequence, it is this: a Visa debit card that spent on-chain USDC at more than 44 million merchants, launched in December 2024 and run by Cypher, is now just a closed programme with a clear cut-off date and a final window to withdraw any leftover balance.

What the card looked like before it closed

A few details from third-party reporting are worth keeping in mind, because they explain what Moonwell was trying to be, even though none of it is the card’s own page any more. Treat the following as unofficial and not confirmed by the card itself.

The card was, in concept, a Visa debit card that let users spend on-chain USDC at more than 44 million merchants (unofficial, per Crypto Briefing, 2026; Diario Bitcoin). It launched in December 2024 (unofficial, per Crypto Briefing, 2026). It was reportedly available in more than 180 countries, excluding New York and Vermont in the US (unofficial, per Crypto Briefing, 2026; Diario Bitcoin). The historical load fee on USDC was around 0.5%, and the foreign exchange fee around 1.5% (unofficial, per the same sources). Apple Pay and Google Pay were supported (unofficial, per Crypto Briefing, 2026). The card also offered a “borrow-and-load” route: borrow against a Moonwell lending position and load the proceeds onto the card, which is a feature few plain-vanilla crypto cards have offered (unofficial, per Crypto Briefing, 2026).

If you are new to Moonwell as a name, none of that helps you today, because none of it is available. The reason it still matters is that it sets the shape of the question “what would replace this?” — a USDC-denominated card, on Visa, with mobile wallet support, ideally with some kind of credit-style flow. The alternatives section below keeps that shape in mind.

If you already have the card

This is the section I would want to read first if I were in your position, so I have put the practical steps up front. If you held a Moonwell card, there are three windows that mattered, and only the third one is still relevant today.

The first window ran until 8 July 2026. That was the last day to load the card, and the last day to apply for a new one. If you missed it, no further loads or new cards were possible.

The second window ran until 8 August 2026 at 00:00 UTC. That was the last day the card actually worked at a merchant, ATM or online checkout. After that, every card was cancelled. If you tried to use a card on 9 August, it would simply decline.

The third window ran until 6 September 2026. That was the deadline to withdraw any remaining balance from the Cypher app or dApp at no fee, and the day the app itself went offline. After that, support ran for 60 more days, ending around early November 2026, but free withdrawals were no longer available.

The infrastructure provider, Cypher (CypherD Wallet Inc.), described the card as a non-custodial wallet: the user held the private keys, not Cypher. That detail matters in practice, because the closure was a programme closure, not a custodial collapse. Rewards earned before 16 July 2026 were honoured, but only if you claimed them before the end date; after that, they were not. Cypher support, at [email protected], was the contact point through 6 September 2026, and through the 60-day transitional period after.

If you are reading this in late 2026 or later and you still have a balance sitting in a Cypher-linked wallet, your path is to look at whether the wallet is non-custodial, in which case your seed phrase and your keys still control the funds, and the underlying tokens are recoverable on-chain. If the wallet was custodial in your case, the support window is the only route, and it is now closed. I have not seen any extension announced.

One thing the Cypher notice does say is that anti-money-laundering law requires it to keep identification and transaction records for at least five years after account closure. That is normal for regulated wallets, and it explains why the support team is no longer reachable, not why your records are gone.

What can go wrong

A closed programme has its own set of risks, and they are not the usual “did I pay too much in fees” questions. They are about timing, about unclaimed rewards, about getting your money out, and about the small but real chance of scams that follow any closure announcement.

The first risk is missing the free withdrawal window. Cypher explicitly said withdrawals were free only until 6 September 2026, and after that the support team carried on for 60 days. If you had any balance left in the card or in the Cypher wallet after 6 September, and you had to use support rather than the app, the fee structure was no longer zero. The exact post-6 September fee is not stated publicly, and the Cypher support team is no longer reachable. I would not assume any refund is still possible.

The second risk is unclaimed rewards. The final rewards cycle covered 2 to 16 July 2026, and the notice said rewards earned before 16 July 2026 would be honoured only if claimed before the end date. “End date” here means the closure date, not a vague “later.” If you earned a small NEXO-style, USDC-style or points-style reward and did not click claim, treat it as gone.

The third risk is confusing the card closure with the underlying wallet. Cypher describes its wallet as non-custodial, which means the user keeps the private keys. That is genuinely different from a custodial wallet, where the provider holds the credentials and you must trust them completely. The closure of the card programme is not the same as a custodial failure: the chain underneath is intact, and if you still hold the seed phrase, you can in principle restore the wallet with another client. The risk is that you do not have the seed phrase, that you have never set it aside, or that you have handed the keys to a third party at some point. Cryptocurrency transfers are typically irreversible, and a lost seed phrase has no recovery path on the network itself.

The fourth risk is the standard one that follows any high-profile crypto closure: phishing. After a card programme ends, the obvious scams are emails pretending to be from Cypher, Moonwell, or Nium, with “final withdrawal” links, “claim your rewards” buttons, and “your account will be deleted” copy. The official support address was [email protected] and the in-app channels are gone. Treat any link you receive in a text or an email as suspicious by default, and use the address you already have, not the one in the message.

The fifth risk, more of a structural one, is regulatory record-keeping. Cypher’s notice says it is required to keep identification and transaction records for at least five years after account closure under anti-money-laundering law. That is normal, but it means your data is still on file somewhere, and you do not have a “delete my account” button to press.

A small point on the company side: Cypher was acquired by Nium, which is a real global payments company, but the details of that acquisition, including the timing and the structure, are not in the public Cypher notice I read. Moonwell’s own company registration, where it is incorporated, and who its principals are, are not stated in the card’s materials. I would not lean on either as a fact I can verify.

I did not find any Trustpilot or Reddit complaints specifically about the Moonwell card, as of 20 October 2026. That is not the same as saying the card was perfect; it is just a reflection of how few cardholders there were, and of how fast the closure came.

What to take instead

Because the Moonwell card is closed, the more useful question is “what would replace it for someone in the same situation?” The honest answer is that none of the alternatives is a perfect drop-in, and each one has its own country rules, fee structure, and stablecoin support. Below are three cards I have looked at that come closest in spirit: a USDC- or stablecoin-leaning card, with mobile wallet support, and a clear application path.

I have not used any of these cards. The notes that follow are based on the cards’ own pages and, where I have flagged it, secondary reporting.

Nexo

Nexo’s card is a Mastercard, issued by DiPocket UAB, available as a virtual card that activates in the app and a physical card that is shipped free. It is offered to residents and citizens of selected European countries, including the EEA and the United Kingdom, and is not available in certain other jurisdictions; the page does not name them. Cointelegraph reported in 2025 that Nexo’s Sofia office was raided by Bulgarian police over alleged Russia-sanctions and AML violations, and I did not find a Nexo statement on this matter specifically about Russian users.

The card has no monthly, annual, or inactivity fee. The virtual card needs at least $50 in the account to activate, and the physical card requires a balance above $5,000 in digital assets and at least a Gold tier. Foreign exchange fees are 0.2% on weekdays in the EEA, UK, and Switzerland, and 2% elsewhere; on weekends those figures rise to 0.7% and 2.5%. Cashback is paid in NEXO tokens or Bitcoin, varies by tier (Platinum 2%/0.5%, Gold 1%/0.3%, Silver 0.7%/0.2%, Base 0.5%/0.1%), and applies only in Credit Mode, with a portfolio of more than $5,000 in digital assets. The card spends from a Nexo account that supports over 100 digital assets, and in Debit Mode it can use stablecoins, crypto, and FiatX assets such as EURx, USDx, and GBPx.

Score: 6.3 / 10. The card is available, well-documented, and cheap to use inside Europe. The catch is geographic: if you are not in the EEA, UK, or Switzerland, the FX fees jump and the cashback thresholds are higher.

Pay3

Pay3 offers virtual, physical, and Apple Pay/Google Pay cards, under license from Visa. Card issuance starts at $5, and the card spends from $0.15 per transaction. The card is collateral-based, meaning a user-supplied collateral sets the card’s spending limit, and the funding side accepts stablecoins or fiat into a multi-currency account.

The card is open to users outside “Restricted Locations” — countries under trade sanctions, export controls, or crypto or fintech prohibitions. The list is not published and may change without notice. Card terms require users to attest they are not a US citizen and not OFAC-blocked. Currency conversion fees start at 1.15%, with foreign exchange, cross-border, and foreign transaction fees of up to 3%. Rewards are not stated.

Score: 6.3 / 10. Pay3 is the most crypto-native in its design: collateral-based, multi-currency, mobile-wallet-friendly. The downside is the fee ceiling, which can run to 3% in non-USD spend, and the unpublished restricted-country list, which means a quick eligibility check is essential.

Exa

Exa is a virtual Visa Signature credit card, accepted at over 150 million merchants in more than 150 countries, with the issuing geography in 45 countries (including Argentina, Brazil, Japan, Kenya, Mexico, Singapore, South Africa, and the UAE, with Nigeria conditional) and 31 US states. Eighteen countries are ineligible, including Belarus, China, India, Iran, Israel, Turkey, Ukraine, Vietnam, and Russia. Russian residents cannot be issued a card, and spending is blocked in Cuba, Iran, North Korea, Russia, Syria, Ukraine, and Venezuela. The rule is by residence, not citizenship.

The card is marketed as free, with no amounts on the page; the help centre has an article on “Exa Card associated fees and costs” that I have not read in full. Funding comes from account collateral in the self-custody Exa App, and the card spends either via “Pay Now” against collateral or via installments (a credit-line flow). Specific coins and networks are not stated, and the help centre lists Visa Signature benefits, a 20% Airalo eSIM discount, and a US activation reward (spend $100, get $20 in XRP via Uphold). Rewards: no cashback stated; “Earn while you spend” refers to yield on the underlying assets.

Score: 6.2 / 10. Exa’s strength is its credit-style flow against self-custody collateral, which is rare in the category. Its weakness is that the fee details and the precise funding rules are not on the main page, and eligibility rules exclude a long list of countries including Russia.

A quick comparison

Nexo Pay3 Exa
Network Mastercard Visa Visa
Form Virtual + physical Virtual + physical + Apple/Google Pay Virtual
Funding 100+ assets; FiatX in Debit Mode Stablecoins, fiat into multi-currency account Self-custody Exa App collateral
FX fee 0.2% weekdays EEA/UK/CH, 2% elsewhere From 1.15%, up to 3% Not stated on main page
Rewards NEXO or BTC, by tier, Credit Mode only Not stated No cashback; yield on assets
Main catch EEA/UK/CH focus Unpublished restricted-country list Long list of ineligible countries

None of these is a USDC-on-Visa, Apple Pay-supporting, borrow-and-load card with a 0.5% load fee. If that exact combination is what you came for, there is no current direct replacement. If you are happy to relax one of those constraints, one of the three above should fit.

Who I would recommend it to

A closed programme is, by definition, not for anyone. So instead of telling you who to recommend it to, let me tell you who this page is genuinely for, and who should just close the tab.

You should keep reading if you held a Moonwell card and want to confirm the closure dates, the withdrawal window, the rewards-claim window, and the support email. The dates are in the section above; the email was [email protected]; the support line ran through 6 September 2026 and then a 60-day tail. I have written that section to be the one you can copy a date out of.

You should also keep reading if you are weighing a USDC-on-a-card setup today, and you want to know what the Moonwell card’s design was, what made it unusual (the borrow-and-load against Moonwell lending positions, in particular), and which current cards come closest. The alternatives section above is your starting point, not your final answer: each of those cards has its own country list, its own fee structure, and its own quirks, and you should read their own pages before you apply.

If you were hoping to find a “how to apply” section, there is not one, and there cannot be one, because there is no application. If you were hoping to find a way to get the old card working again, there is not one of those either: every card was cancelled on 8 August 2026. The honest answer to both questions is to pick one of the alternatives and apply there.

Finally, if you are reading this because you are not sure whether you should be doing anything with a balance you forgot about, the answer is the same as for any non-custodial wallet: check whether you have the seed phrase. If you do, you can in principle restore the wallet with another client and move the funds on-chain yourself. If you do not, the funds are reachable only through the now-closed support line, and that line is no longer answering.

Frequently asked questions

What is the Moonwell card?

It was a Visa debit card that let users spend on-chain USDC at more than 44 million merchants, launched in December 2024. The card was operated by the infrastructure provider Cypher. It is no longer being issued, and existing cards were cancelled on 8 August 2026.

Is the Moonwell card still available?

No. Card loading stopped on 8 July 2026 and new issuance ended on the same day. The last day to use an existing card was 8 August 2026 at 00:00 UTC. The app, dApp, and free withdrawals closed on 6 September 2026.

How does the Moonwell card work?

The card drew on a pre-loaded balance. Users loaded USDC on-chain, or borrowed against Moonwell lending positions and loaded the proceeds, and the card then spent that balance at merchants via the Visa network. Apple Pay and Google Pay were supported, both unofficial. None of this is operational today.

Is Moonwell a bank?

No. Moonwell is the brand on the front of the card, and Cypher (now owned by Nium) was the infrastructure provider. Banks and electronic money institutions are the entities that actually issue cards under a network’s licence. Moonwell’s own company registration and principals are not stated in the card’s materials.

Who owns Moonwell?

Moonwell’s ownership is not stated in the card’s public materials. The card’s infrastructure provider, Cypher, was acquired by Nium, and the acquisition was announced on 8 July 2026. The structure of that deal, and Moonwell’s relationship to it, is not described in the public Cypher notice I read.

Is the Moonwell card available in my country?

For practical purposes, no: no new cards are being issued anywhere. Historically, the card was reportedly available in more than 180 countries, excluding New York and Vermont, but those figures are unofficial and the programme is closed. If you are choosing a card today, you should check the country list of one of the alternatives in this page.

How did you top up the Moonwell card?

Card loading was via on-chain USDC, or via a “borrow-and-load” flow that borrowed against a Moonwell lending position and loaded the proceeds onto the card. Both are unofficial and both stopped working on 8 July 2026, when card loading ended.

How do you withdraw money from Moonwell now?

The free withdrawal window ran until 6 September 2026 via the Cypher app or dApp. After that, Cypher support ran for 60 more days. If you still hold the seed phrase to a non-custodial wallet, you can in principle restore the wallet with another client and move the funds on-chain yourself. If the wallet in your case was custodial, the only contact path was Cypher support, and that line is now closed.

How do you add the Moonwell card to Apple Wallet?

The card supported Apple Pay, per unofficial reporting, but adding a card that has been cancelled is not useful. The Apple Pay token would have been removed or made inactive along with the card on 8 August 2026, and the app that issued the token is offline.

Is the Moonwell card safe?

The card’s infrastructure was a non-custodial wallet, which means the user kept the private keys, and that is genuinely different from a custodial wallet. Safety depends on whether you held the seed phrase and whether you keep it safe. After 6 September 2026 the support line is closed, so any post-closure issue has no clear contact path. Treat any “final withdrawal” link you receive by email or text as a phishing attempt by default.

Content created by the «cryptocardoptions.com» team