Ratings · Measured, not sponsored
Crypto cards
We set out expecting to find that every crypto card is a bad deal. Two of them are genuinely good — and they are good for the same reason: they refuse to sell you a token. Here is the ranking, and the arithmetic behind it.
Up to 4% back in Bitcoin on a real credit line from a real bank. No token, no lock-up, no annual fee, no FX fee, no conversion spread. Best for a US spender by a distance.
The same idea on Amex — 2–4% back in Bitcoin, no proprietary token. Nets positive after the Coinbase One membership.
Where a card sells the same cashback for cash or for a token stake, take the cash. Crypto.com and Wirex priced it themselves, and the cash won every time.
What we did · July 15, 2026We read each issuer’s published terms, priced every staking token against its real 12-month return, and worked the net reward on $24,000 of annual spending — then checked who actually issues the card and whether an American can get it.
Why the headline rate is not the reward
The staking ladders invert as you climb them. Cashback is linear in what you spend; token risk is linear in what you stake — and the stake grows a hundredfold from tier to tier while your spending stays flat. So the top tier is always the worst deal, and it always looks like the best one.
This is why the ranking below leads with the cards that pay a flat rate and stake nothing — and why, for a US reader, the good answers turn out to be the ones that do not ask you to hold a coin at all.
The ranking
Click any row for the quick read, or open a company for its full profile and per-criterion scores. Measured July 15, 2026. How we rate.
HOW WE SCORED THIS — CRITERIA AND WEIGHTS+
A crypto card is worth having only if the reward survives contact with the fine print. Most of the headline cashback here is gated behind staking the issuer’s own token — and once you price that token honestly, the reward often turns negative. So we rank on the reward you actually keep, on whether the real cost is disclosed, and on whether an American can even get the card.
Not the headline percentage — the reward net of what earning it costs. Where the rate is gated behind staking a volatile issuer token, we price that stake against the token’s real drawdown and ask whether a cash alternative exists. A card that pays a flat rate in Bitcoin with nothing to lock up keeps its whole number; a card whose 8% requires locking a token down 96% from its high does not.
The cost that touches every purchase is the crypto-to-fiat conversion spread at the till, and almost nobody states it as a number. We check whether the spread and the foreign-exchange markup are published. Two exchanges disclose more than the US market leader; several disclose nothing and say only that charges “may apply”.
Who actually issues the card — the real bank or e-money institution behind the brand — and what happens to your balance if the program stops. Program suspensions are not hypothetical: the 2020 Wirecard collapse froze several of these cards overnight, and BlockFi’s card died with its bankruptcy. A real, named, regulated issuer scores; an opaque one does not.
Whether an American can actually get the card, and in which states. Several of the highest-headline cards are unavailable in the US entirely, which for our audience settles the matter regardless of the rate.
The honest arithmetic on a realistic $24,000/year of spending, from the issuer’s own published terms and public token prices — including the opportunity cost of any locked capital and the token’s realised return. Where the net is positive we say so; where it is negative we show the working.
The app, Apple and Google Pay support, virtual cards, and ATM access — the ordinary things that make a card usable.
Weights sum to 100. If we cannot verify a criterion, we delete it rather than score it on impressions — read the methodology.
Is there a free alternative?
The best card here asks you to stake nothing, and it is not close.
The Gemini card pays up to 4% back in Bitcoin on a normal credit line from a real FDIC-member bank, with no token, no lock-up, no annual fee and no foreign-transaction fee. It wins precisely because it refuses to sell you a coin. The one rule that survives this entire category: never buy the issuer’s token to unlock a cashback rate. If a card offers the same rate for a cash subscription or a token stake — as Crypto.com and Wirex both do — take the cash every time. They priced it themselves, and the cash option wins in every case.
What changed since last time
- 2026-07-15Category published as a ranked rating.An earlier version refused to rank these on the grounds that the stake-gated ones are bad. That skipped the real finding: some crypto cards are genuinely good, and a reader deserves to be told which.
Questions
What is the best crypto card for an American?+
The Gemini Credit Card, and it is not close. It pays up to 4% back in Bitcoin on an ordinary credit line from a real FDIC-member bank (WebBank), with no token to stake, no lock-up, no annual fee and no foreign-transaction fee. Because you spend dollars and receive crypto as a reward, no crypto is sold at the till, so there is no hidden conversion spread. The Coinbase One card is a close second on the same principle. Both are good precisely because they refuse to sell you a token.
Why do the cards with the highest cashback rank lowest?+
Because their headline rate is gated behind staking the issuer’s own coin, and every one of those coins is down over the past year — CRO about 48%, WXT 69%, PLU 80%. Price the stake honestly and the reward inverts. Wirex’s 8% requires locking 7.5 million WXT, worth about $29,938 a year ago and about $9,179 now: the 8% earned roughly $1,920, and the stake lost about $20,759. The same plan pays 4% with no token locked, and that nets +$600. The token did not raise your return; it cost you about $21,000 to capture $960 of extra cashback.
Is there a rule that survives the whole category?+
Yes: never buy the issuer’s token to unlock a cashback rate. You do not need to trust our arithmetic to follow it — just take the issuer’s own cash offer instead of its token offer. Crypto.com sells the same 3% for a $29.99/month subscription (nets +$360) or a $5,000 CRO lock (nets −$1,884). Wirex sells 4% for cash (+$600) or 8% for 7.5m WXT (−$20,396). They priced their own token against their own subscription, and the cash won every time.
What about the conversion fee I never see?+
It is the one cost that touches every purchase, and Crypto.com, Wirex, Plutus, Bitpanda and Uphold do not publish it as a number — Crypto.com’s own page says only that "interbank exchange rates, funding fees, and other charges may apply". At an estimated 0.5–1.5%, that invisible spread on $24,000 of spending is $120–$360 a year, enough to halve a 2% cashback. The cards that avoid it entirely are the credit cards (Gemini, Coinbase One), where you spend dollars and no crypto is converted at all. Among the rest, only Bybit and Nexo — neither available in the US — actually publish the number.
Can the card just get cancelled?+
It happens. In 2020 the Wirecard collapse froze the Crypto.com, Wirex and TenX cards overnight; Binance closed its European card outright in December 2023; and BlockFi’s card died with its bankruptcy, where its custodial users became unsecured creditors and actually lost money. Bitpanda even retired BEST — the token it had told cardholders to accumulate. So issuer and program safety is a real criterion, which is why a card from a named, regulated bank scores above one from an opaque issuer even at the same cashback rate.
How this is funded
It is not. There are no affiliate links on this page or anywhere on this site, no paid placements, and no sponsored positions. Nobody in this table can buy a place in it, accelerate their inclusion, or influence a score — and none of them paid us anything, because there is nothing here to buy. The full policy.