WEEKLY WRAP · 2026-09-22
5 quick hits · market context · 43 cards updated
SHUTDOWN
Friday's rotating check went through 50 cards and changed the status of two. Buenbit is the one worth sitting with. Argentina's Buenbit was bought by Nexo in December 2025, accounts migrated between April 1 and May 3 this year, anything left behind was swept automatically by May 8, and the Buenbit account was deactivated after that. Type buenbit.com today and you get the Nexo Argentina site. The 1% to 5% cashback ladder we had on file, tiered on your previous 30 days of fiat-to-crypto volume and paid in DAI, BTC, ETH, USDC or nuARS, does not exist anywhere any more. We carried that card as active until Friday. Four months. The thin defence is that the acquisition was announced as a migration rather than a shutdown, Nexo's own Argentina dual debit and credit launch in July got written up as a launch, and nobody files a story saying a card quietly stopped existing in the spring. The second one is OnilX, which we flagged to you on September 15 as marked active and unusable. Now the company has said it itself: OnilX is concentrating on B2B infrastructure, asset withdrawals opened from August 17, onilx.com.br is an institutional payments and custody page, and the card URL returns a 404. That is the second independent confirmation our rules were waiting for, so last week's flag is this week's status change. If you still have money there, read this part twice. Withdrawals are reported to require signing a broad, general and irrevocable release before the funds arrive, and that sits on top of a court order freezing company assets and 48 active Reclame Aqui complaints. Our read: a card dies in one of two ways. Loudly, with a wind-down schedule and a balance-withdrawal deadline, which is what Cypher did and what we caught in a week. Or silently, absorbed into an acquirer, where the app keeps opening for a while and then one day redirects. The second kind is the one every comparison site on the internet is currently getting wrong, ours included.
Source: Buenbit Buenbit Card OnilX Crypto Card Nexo Card →
RATES
Eight days from now the Intermediary Cashback Programme stops. It has run since November 9, 2025, funded by Gnosis Ltd, and it pays 1% if you hold at least 0.1 GNO in your Gnosis Pay Safe, 2% at 1 GNO, 3% at 10, 4% at 100, with a 1% OG NFT bonus taking the ceiling to 5%. The GNO is held, not staked and not locked, payouts land weekly, and your weekly spend cap scales with your holdings. What comes next is the part worth planning around. Gnosis is not renewing its own programme. Partner-led incentives take over instead, Zeal for Europe and PicnicBR for Brazil, and the Brazilian one is already running. Nobody has published what those pay. Our read: a first-party cashback programme and a partner-funded one are different products wearing similar badges. The first is an issuer paying you to hold its token, and it lasts as long as the issuer wants distribution. The second is a third party buying its way in front of you, on its own terms, for as long as its budget holds. Neither is worse on its face. They just fail differently, and only one of them is a number Gnosis controls. We should also own that we have had this date wrong before, in the opposite direction: until August 14 our entry said the programme had been discontinued on June 30, and it had not been. The correction is in the note and has been since. If you are holding 100 GNO specifically for that 4%, October is when you re-run the arithmetic, because right now there is no published replacement rate to run it against.
Source: Card Pilled Gnosis Pay →
RATES
For months our Kardpay entry recorded the cashback tiers as not published, with a noted contradiction between Kardpay's own "up to 2%" marketing and the 4% and 8% figures third-party writeups keep repeating. The site now publishes the actual ladder and it settles the argument in Kardpay's favour: 1% at a $3,000 KDY stake, 1.5% at $6,000, 2% at $18,000, each locked for 180 days after activation. There is no rate at all without a stake. So the widely circulated top rate is four times the real one, and the real one asks you to lock eighteen thousand dollars of a card issuer's own token for half a year to earn it. Two more things came out of the same read. The product tiers are Core at $29, Pro at $99 and Elite at $299 with one-year validity, not the two-year validity we had recorded, and top-up runs 5%, 4% and 3% by tier rather than the flat 4% in our old note. And Kardpay writes its thresholds as "$3,000 KDY", using dollar notation for a token amount, without saying anywhere whether it means 3,000 tokens or three thousand dollars' worth of them. Those are very different asks. We recorded the ambiguity rather than picking the flattering reading. Here is the uncomfortable bit, and it is about our tooling rather than Kardpay's. We run an audit that flags vague tier requirements, and it never flagged this card, because the unresolved tier had been recorded at 0% and the audit scores rates rather than the absence of one. A fabricated or unknown tier parked at zero is invisible to it. That is the second time this month a number looked settled because our own machinery had no way to complain about it.
EXPANSION
Fasset now publishes a named eligibility list, and it is nothing like the shape we had. Thirty-six countries: nine in Asia-Pacific (Australia, Bangladesh, Hong Kong, Japan, Malaysia, New Zealand, the Philippines, Singapore and Pakistan), nine in Africa (Egypt, Ghana, Kenya, Morocco, South Africa, Uganda, Zambia, Côte d'Ivoire and Senegal) and eighteen Americas and Caribbean markets. Against that sits an exclusion list that blocks issuance to residents of the UAE, Saudi Arabia, Bahrain, Qatar, Oman, Kuwait, Indonesia, Nepal, Nigeria, India, Vietnam, Turkey and all of mainland South America. Read those two lists together and the card is not a MENA product and not a South Asian one either, outside Pakistan and Bangladesh. Every Gulf state is out. Fasset's UAE and Indonesia regulatory approvals are real, and they cover the platform rather than card issuance, which is the same distinction that had us recording Turkey as available back in July. This is the third correction to this entry since then. We had 125-plus countries, which was the platform figure. On September 6 we cut that to the 40-plus its card page stated. Now there is an actual list and it is 36. While we were in there, the other thing this entry has been carrying: Fasset launched in April 2026 in partnership with Tether, marketed as the first gold-backed stablecoin neobanking card, and the words gold, XAU and cashback appear nowhere on the current card page or in the cardholder terms. The live mechanic is OWN Points earned on spend with no published conversion rate to money, which is why we score it at 0% rather than guessing. Our read: a launch press release is a marketing document with a shelf life, and five months later it is describing a product that may not be the one shipping.
Source: Card Pilled Fasset Card →
KYC
Somebody circulated a graphic listing 16 no-KYC crypto cards. We already tracked 12 of them, which made it cheap to check, so we did. Hyperbeat came out of it as a genuine addition and its help centre publishes the whole model, including the identity verification process: government ID plus a liveness check. Kardpay, which shows up on lists like this constantly, opens with connect-wallet and then asks for phone number and country. Neither is a no-KYC card. Since the entry is new, the rates: 0.2% on all card spend with no deposit and no minimum, plus four separate +0.2% boosters for holding $1,000, holding $10,000, spending $1,000 in 30 days and spending $10,000 in 30 days, which stack to 1% if you do all four. Above that is a trading ladder that pays 3% at $1M of 30-day volume, 5% at $5M, 8% at $10M, 10% at $25M and 12% at $50M or more, each capped by a spend ceiling, so the 12% covers about $41,700 of card spend and nothing beyond it. Everything pays in WHYPE within 48 hours of the transaction clearing. The card is a Visa Signature with two modes, and the second one deserves attention: Cash Mode spends your USD balance with no interest, while Credit Mode borrows at a variable APR with no grace period against collateral held in your Hyperbeat account at 40% to 75% LTV, liquidating at 62.5%. Issuer is Third National, operator is Zoeion Ltd. FX runs 0.4% to 1%, ATM is $1 plus 0.65%. Haven, Fasqon and CryptoCash came off the same graphic and failed our new-card checklist, so they sit in the backlog rather than the database. Our read: "no KYC" in this market mostly means KYC you have not triggered yet, and a graphic is not a source. Adding Hyperbeat also surfaced a bug on our own card pages, which is that the cashback FAQ listed every tier straight after the headline rate, so the answer read "offers 0.2-1% WHYPE cashback. There are 8 tiers ... Trading 50M+ (12%)", contradicting itself inside one sentence and inside the FAQ structured data Google reads. Tiers above the headline now get their own sentence. Checked against every card page on the site, exactly two changed: Hyperbeat's and Trade Republic's, which has the same shape with a 1% headline and a 2% Crypto Saveback tier.
Source: Hyperbeat Hyperbeat Card → Kardpay → Trade Republic Card →
Market Context
Almost nothing happened this week. No launches, no shutdowns, no regulatory events, no rate moves: our news scans came back three times running with essentially nothing dated inside the window. And yet four cards in our database had a material fact change, two of them changed status, and one of them had been dead since May. Notice where the errors were. Not in the news. In the records. Buenbit's cashback ladder had been gone for four months while every comparison site, ours very much included, kept publishing 1% to 5%. Kardpay's real ceiling is 2% and the internet says 8%. Fasset was filed as a MENA card while blocking issuance to every single Gulf state. A widely shared graphic called Hyperbeat and Kardpay no-KYC cards when both run identity checks and publish that they do. Our read: this corner of the internet has a copying problem, and the copying is mostly of each other rather than of providers. A number gets into circulation from a launch announcement, gets republished, gets republished again with the source now two hops back, and by the time anyone opens the provider's help centre the figure has the weight of consensus behind it and none of the evidence. The only fix is unglamorous. Open the page. Read the terms. Write down what you could not confirm instead of rounding it to what the last aggregator said. We did that four times this week and were wrong four times, which is roughly the hit rate we expect and the reason we keep publishing the misses. On the numbers: 204 cards, 168 active, 31 discontinued, 4 announced and one, Anytap, that we genuinely cannot classify, which is a category we had to create this week and will probably need again. Three cards added, forty entries re-checked and roughly a dozen of those actually corrected. If you want a single thing to act on before October, it is Gnosis Pay: the 5% ceiling has eight days left and there is no published number waiting behind it.
Cards updated this week: Buenbit Card OnilX Crypto Card Kardpay → Fasset Card → Fuse Card → Pinto Pay → Laso Finance Card → X Money Card → UTORG Card → Pera Card → Satoshi Tango Card → Lemon Card → Plenti Card → Eurasian Bank Crypto Card → Binance Card (LATAM/Africa) → Binance Japan Card → Busha Card → Kredete Card → SPENDL Money Card → LuqasCard → HashKey Credit Card → HashPort Card → UPay Card → bitFlyer Credit Card → MiniPay Card → Stablecard by Western Union → Lemon Card Crédito → Decrypto Card → LB Finanzas Card → Trustee Plus Card → DPT Card → Goblin Card → Agora Card → Spritz Card → KOSH Global Card → Keytom Card → Bitrefill Card → Karta Card → Plutus Card → MoonAgents Card →
New cards added: Hyperbeat Card → Fizen Card → Anytap Card →
← Previous issue: September 8–14, 2026 | All issues: Weekly Wraps archive →