Visa and Bridge push stablecoin cards toward 100 countries

Visa and Bridge, the stablecoin infrastructure company Stripe bought in early 2025, are expanding their card programme from 18 countries to a planned 100-plus across Europe, Asia Pacific, Africa and the Middle East by the end of the year. The announcement is short. The machinery behind it is the most complete picture yet of how stablecoins and card networks actually fit together, and it is worth taking apart properly.

What the product is

The consumer-facing half is simple. A developer, a wallet company, a fintech, an exchange, uses Bridge's APIs to issue a Visa card whose funding source is a stablecoin balance. The cardholder taps at any of Visa's roughly 175 million merchant acceptance points, and from the merchant's side nothing unusual has happened at all: an authorisation arrives, a settlement follows, in local currency, on the same schedule as every other Visa transaction.

The interesting work happens in the half-second in between. When the authorisation request lands, the stablecoin balance has to become spendable fiat: a conversion executed at authorisation time, so the merchant is never exposed to the token and the cardholder never pre-funds a fiat account. Bridge operates that conversion layer, alongside the custody, compliance and reporting that a card programme drags with it.

Two crypto wallets, Phantom and MetaMask, are already live on the rails. That detail deserves more attention than it usually gets: wallets are where stablecoin balances actually sit, and a card that attaches directly to the wallet skips the step, moving funds to an exchange or a bank, where most consumer crypto spending has historically died.

A short history of the crypto card

The industry has run this experiment before, in three generations that each taught an expensive lesson.

The first generation ended abruptly in January 2018, when Visa terminated the European issuer whose licence sat under most early crypto debit programmes and thousands of cards died over a weekend. The lesson: the BIN sponsor is a single point of failure, and programmes borrowed against marginal issuers are borrowed time.

The second generation, the exchange era, put cards on sounder licences. Crypto.com spent lavishly on its metal-card programme, and Coinbase shipped a card in Britain in 2019 and America in 2020. These proved demand but shared a design constraint: they spent from custodial exchange balances, so the user's assets sat with the platform, and the card business inherited every risk of the exchange running it.

The third generation put rewards ahead of prudence. BlockFi's bitcoin-rewards credit card, launched in 2021, was genuinely popular right up until its issuer followed the 2022 credit collapse into bankruptcy, taking the programme with it.

The current generation, of which the Bridge-Visa stack is the fullest expression, is a response to all three failures: regulated, purpose-built sponsors rather than marginal ones; funding from stablecoin balances rather than volatile assets; and, in the newest designs, custody that stays with the user until the moment of authorisation.

The anatomy underneath

A card programme is never just two companies. Underneath this one sits the standard anatomy: a BIN sponsor, the regulated bank whose licence the cards are issued against; a programme manager handling the operational load; the network, Visa, moving authorisations and clearing files; and now a stablecoin layer where the deposit account used to be.

That is where Lead Bank comes in, and where the announcement gets structurally interesting. Through Bridge's partnership with the Kansas City-based bank, transactions from these programmes can settle with Visa on-chain, as part of the network's wider stablecoin settlement pilot. Instead of the issuer's bank wiring fiat to Visa through conventional correspondent channels on banking days, the settlement obligation itself moves as tokens over a blockchain.

Visa has been building toward this for five years. It first piloted stablecoin settlement in 2021, letting a crypto issuer settle obligations in USDC rather than wiring dollars, and has expanded the programme in stages since, adding chains and counterparties. The Bridge arrangement extends the idea from crypto-native issuers to a whole class of card programmes, with a regulated US bank in the loop. Visa says it is also evaluating support for Bridge-issued stablecoins as a settlement pathway in their own right, which would close the circle entirely: a card programme funded in a stablecoin, settling with the network in the same asset.

Cuy Sheffield, Visa's head of crypto, compressed the strategy into one sentence: the company is committed to meeting businesses where they operate, "and increasingly, that's onchain."

Why the country count is the hard part

Eighteen countries is not a technology boundary. It is a licensing map. Card issuance touches banking regulation in every market where the card is issued, and attaching a crypto asset as the funding source adds a second regulatory layer that most jurisdictions have only recently defined at all.

The expansion targets, Europe, Asia Pacific, Africa and the Middle East, map to where those definitions have recently arrived. MiCA gave the European Union a unified framework for stablecoin activity. Licensing regimes in the Gulf, Singapore and Hong Kong did the same for their markets. The jump from 18 countries to 100 is a bet that the regulatory doors that opened between 2024 and 2026 stay open, and that Bridge can assemble the local sponsor and licensing relationships behind each one fast enough to matter.

The prize for being first is durable. Card programmes are sticky: once a wallet has issued cards against one infrastructure provider, migrating means reissuing plastic, renegotiating a BIN sponsor and re-clearing compliance in every market. Whoever holds the issuing relationships in these hundred countries when they open will be very hard to displace, which explains the pace.

The economics, in numbers

Card economics are unforgiving arithmetic, and stablecoin funding changes several terms at once.

On the revenue side, interchange in the United States runs meaningfully north of one percent on credit programmes, while regulated debit is capped near a fixed cents-per-transaction level, which is why most crypto programmes are structured as prepaid or credit rather than bank debit. That interchange pool is what funds cashback, and it is why rewards-led crypto cards keep appearing: the margin exists, if the programme can control its other costs.

On the cost side, a stablecoin-funded programme replaces one expense and adds another. It removes the cross-border FX spread that conventional cards charge when a cardholder spends across currencies, often around three percent, which is exactly the fee that makes cards expensive in emerging markets. It adds a token-to-fiat conversion at authorisation, whose spread is a fraction of that when the token is a dollar instrument spending against dollar pricing. Net, for a user in São Paulo or Lagos holding dollars in a wallet, the stablecoin card is frequently the cheapest dollar-spending instrument available to them, which is the quiet engine under all of these launches.

Why the demand sits in emerging markets

That points at the real geography of the 100-country plan. In the United States and the euro area, a stablecoin card competes with excellent domestic alternatives and wins mainly on novelty. In markets with weak currencies, capital controls or thin card penetration, it competes with informal dollarisation, and wins on safety and acceptance. The corridors where stablecoin supply already circulates hardest, Latin America, Africa, South and Southeast Asia, are precisely the regions named in the expansion. The product is not chasing crypto enthusiasts. It is chasing the several billion people for whom a spendable dollar balance is the upgrade.

The Stripe angle

Bridge was a striking acquisition when Stripe made it, reportedly its largest ever, and this programme is the clearest public evidence of what it was for. Stripe processes conventional card payments at global scale; Bridge gives it the token layer: issuance, custody, conversion and now card funding, as an API product line. The pitch to a business launching its own stablecoin is that the token need not be an island. Through Bridge it can sit inside a card programme from day one, spendable at every Visa terminal on earth, and Bridge chief executive Zach Abrams made exactly that point in the announcement: businesses launching custom stablecoins can use them "seamlessly within card programs."

That pitch lands in the middle of a crowded field. Rain and other issuing platforms are building the same stack for stablecoin-collateralised programmes; Coinbase has run a card for years; Ethena launched a self-custodial neobank card this week; and the card networks themselves are hedging every bet, with Mastercard running parallel stablecoin pilots and both networks backing the Open Standard consortium. Nobody in the payments industry doubts any longer that stablecoin balances will fund card spend. The open question is whose infrastructure carries it.

The frictions that remain

Three are worth naming, because the announcement does not.

Economics: interchange on card transactions is what funds rewards and platform margins, and a conversion layer adds cost inside a fee pool that is already contested in every regulated market. Whether stablecoin-funded programmes can sustain competitive rewards at scale is unproven outside crypto-enthusiast niches.

Consumer protection: card schemes carry chargeback and dispute rights that do not exist natively on chain. Reconciling a reversible card transaction with an irreversible token conversion is solvable, the conversion provider eats the timing risk, but it concentrates a new kind of exposure in the Bridge layer that will get tested the first time a large programme has a fraud event.

And supervision: a US-sponsored card programme funded from a self-custodied token balance sits at the intersection of banking, payments and the GENIUS Act's new perimeter. The rulemaking published by Treasury this same week will decide how much of this stack counts as offering stablecoins to US persons, and the answer will shape which of these programmes can ever come home to the American market.

What on-chain settlement changes for a bank

The settlement leg deserves one more level of concreteness, because it is where the money is. Card networks settle by netting: at day's end each issuer and acquirer owes or is owed a net position, moved through settlement banks in each currency, funded by nostro balances that sit idle precisely so the money is there at cut-off. Those idle balances are pure cost, and the cut-offs are why nothing settles on a Saturday.

A tokenised settlement leg attacks both. Obligations can move at any hour, so the pre-funding buffer shrinks toward the actual daily flow; and because a stablecoin transfer is final on confirmation, the reconciliation tail, the file-matching that follows every conventional settlement cycle, collapses into the transaction itself. For a large programme, the released working capital is measured in days of volume. That is the number that turns a pilot into a migration, and it is why the phrase settlement optionality appears in these announcements with such regularity: the networks are letting members discover the arithmetic themselves.

The half-second nobody sees

One more piece of anatomy, because the whole product lives inside it. When a cardholder taps, the authorisation must return in well under a second. In that window the stack has to: identify the funding wallet, price the token-to-fiat conversion, reserve or execute it, screen the transaction, and answer the network. Different providers slice it differently, some pre-convert into a spending balance, some convert per transaction and eat the intraday drift, but the design constraint is universal: the blockchain is never allowed to be on the critical path of the tap. Settlement can be on-chain; authorisation cannot wait for it. The engineering that makes a stablecoin card feel like a card is precisely the engineering that hides the stablecoin.

What would prove it

Three observable milestones, none of them press releases. The country counter: whether live markets pass fifty before year-end, which would make the hundred credible. The settlement asset: whether Visa moves from evaluating Bridge-issued stablecoins as a settlement pathway to accepting one, which would be a first. And the American question: whether these programmes, currently notable for their absence from the US market, arrive once Treasury's new issuance rule settles who may offer what to whom. When a Bridge-powered card ships in the fifty states, the regulatory era that began this week will have delivered its first consumer product.

What it means

Strip the branding away and this is the story: the world's largest card network is rebuilding its plumbing so that a token balance is a first-class funding source, and a regulated bank is settling card obligations on a blockchain. The 100-country target will slip or it will not; targets do. The direction has stopped being ambiguous. Cards were the first thing stablecoins were obviously good for at consumer scale, and the infrastructure is now being poured to make that permanent.