Two networks, two different bets. One sells reach, the other sells timing, and the choice decides how a card program funds itself.
Both card networks now accept stablecoins as settlement currency. That is where the similarity stops. Visa has spent eighteen months adding chains and publishing volume. Mastercard spent June rebuilding when settlement happens instead of where it lands.
For an issuer or a program manager choosing between them, the comparison that matters is not which network is more committed. It is which coin the program holds, which chain it can move on, what day of the week the money is due, and whether the program is allowed to settle with the network at all.
What settlement means here, and why timing is the problem
Card settlement is the leg after authorization. A cardholder taps, the merchant gets paid by an acquirer, and at the end of the cycle the network calculates who owes whom. The issuer owes the network, and the network pays the acquirer.
Stablecoin settlement changes the currency of that obligation, not the obligation itself. Instead of wiring dollars through a correspondent bank on a business day, the issuer sends a token on a public chain and the network credits it.
The reason anyone cares is the funding gap underneath. A card program owes the network on the daily settlement cycle, and the money from its customers arrives on a different schedule. Traditional programs cover the difference by parking cash in a prefunded account, which is dead capital sitting there to satisfy a timing mismatch. Every design decision in this story traces back to that gap.
What Visa offers
Visa’s approach has been to widen the menu and let partners pick.
In April the network added Arc, Base, Canton, Polygon and Tempo, taking it to nine chains alongside Avalanche, Ethereum, Solana and Stellar. The coin list dates from the previous summer and runs to four names: USDC, PYUSD, USDG and EURC. That last one makes Visa the only network of the two that can settle an obligation in euros instead of dollars.
Rubail Birwadker, who runs growth products, explained the five blockchains addition as a response to partners building across several networks at once and wanting one settlement layer that spans all of them.
The volume figures are the part competitors quote. Visa disclosed on September 8 that stablecoin settlement had reached a $20 billion annualized run rate, more than fifteen times the level a year earlier, with payment volume up nearly 200 percent and more than 160 stablecoin-linked card programs live in its fiscal second quarter. In April, the same measures were $7 billion and 130 programs across more than 50 countries.
One caveat belongs next to that headline. An annualized run rate extrapolates from recent activity. It is not a statement that Visa has settled $20 billion this year, and anyone building a forecast on it should treat it as a rate of travel rather than a distance covered.
Weekends are the gap in this setup. In August, Visa started a pilot with Nium under the Monetary Authority of Singapore’s BLOOM initiative to test seven-day settlement, covering weekends and public holidays in regulated dollar and euro tokens. It is a pilot in one market, which is a fair description of where Visa’s always-on capability currently sits.
What Mastercard offers
Mastercard went at the calendar instead of the map.
Its June 3 announcement of expanded settlement capabilities added intraday, weekend, and holiday settlement, which is the thing that removes the prefunding requirement instead of repricing it. A program that can settle on a Saturday does not need to hold cash across a weekend to cover Monday’s obligation.
The supported coins are USDC, PYUSD, USDG, USDP, RLUSD and SoFiUSD, across eight chains: Arbitrum, Base, Canton, Ethereum, Polygon, Solana, Tempo and XRPL. ARQ, CBW Bank, Cross River, Lead Bank and Nuvei were named as the first participants, starting in the United States and Latin America.
Raj Dhamodharan, who leads blockchain and digital assets at Mastercard, put the thesis in one line: “The next phase of stablecoin adoption is about real-world utility, especially in settlement, where timing and liquidity matter most.” Jackie Reses of Lead Bank was blunter about why a bank signs up: “The future of financial infrastructure is 24/7, and on-chain settlement is where that future becomes real.”
What Mastercard has not published is a number. No settlement volume, no program count, no run rate. That absence is not evidence of failure, since the capability is three months old, but it does mean any comparison of scale between the two networks rests entirely on one side’s disclosures.
Where the two menus actually differ
Lay the lists next to each other, and the overlap is larger than the marketing suggests.
Six chains appear on both: Base, Canton, Ethereum, Polygon, Solana and Tempo. Visa alone carries Arc, Avalanche, and Stellar. Mastercard alone carries Arbitrum and XRPL. A program already running on any of the six shared networks has a genuine choice of card network. A program built on Stellar or XRPL does not.
The coin lists overlap less, and they run the other way. USDC, PYUSD and USDG are on both. Visa adds only EURC. Mastercard adds USDP, RLUSD and SoFiUSD, giving it six names to Visa’s four.
So the shape of the bet inverts between the two lists. Visa offers more places to settle, Mastercard offers more things to settle in, and each network’s weakness is the other’s pitch.
Two absences matter more than the additions. Neither network settles in USDT, which is the largest stablecoin in circulation, so a program holding it converts before it can pay either network and takes the spread every cycle. And Mastercard has no euro token at all, which leaves a European program carrying currency risk on a dollar coin between authorization and settlement unless it converts first.
Mastercard’s additions point the other way, toward issuers with their own tokens. RLUSD comes from Ripple, SoFiUSD from a bank holding company, USDP from Paxos. That looks less like a liquidity decision and more like an invitation to institutions that want to settle in something they control.
The gate nobody mentions in the announcements
Both networks describe settlement options as though every program can use them. In practice, there is a membership question first.
Settling directly with a card network requires being a licensed participant in it. A principal member holds its own license, its own BIN range, and its own settlement relationship, and can therefore send tokens to the network itself. Most stablecoin card programs are not principal members. They sit underneath a sponsor bank or a program manager that holds the license, and that sponsor settles on their behalf.
So a program’s real question is not which coins its network supports. It is which coins its sponsor will accept, on which chain, and whether the sponsor passes the settlement timing through or holds it. A sponsor that settles with Mastercard on a Saturday but only sweeps its programs on Monday has kept the benefit for itself.
The concentration in that sponsor layer is its own subject, covered in our piece on the card enablers. For choosing a network, the point is narrower. Read the sponsor agreement before reading the network’s press release, because the sponsor agreement is what governs.
The financing layer that arrived with the numbers
The most interesting part of Visa’s September disclosure was not the run rate. It was the credit product published alongside it.
Credit Coop runs a revolving facility denominated in stablecoins and secured only by settlement receivables, meaning the future cardholder payments a program expects, with no crypto collateral behind it. Visa’s authorized daily settlement files are piped to Credit Coop through a secure channel and combined with on-chain history to size facilities and verify repayment.
Repayment runs through a contract Credit Coop calls a Spigot, a programmable lockbox that services interest and tops the line back up before any remaining cash reaches the borrower’s operating account. It is the bank lockbox structure that has financed receivables for decades, rewritten so the collateral and the repayment both move on-chain.
Rain, a Visa principal member, has financed about $2 billion of settlement through the facility since August 2023 across more than 2,000 draws. Credit Coop reports more than $2.5 billion in cumulative volume since 2023 and says it has seen more than 3,000 borrow events and 9,000 repayments with no defaults as of August 19. That default record comes from Credit Coop and not from an audited source, and it deserves the caution any self-reported credit statistic gets. Visa separately says wider lender participation has cut borrowing costs for participating programs by as much as 30 percent, without publishing rates or a method.
Karta, another stablecoin card program, used the facility before raising a $125 million institutional credit line, which suggests the product works as a first rung more than a permanent home. Rain holds principal membership at both networks, so it is one of the few programs able to test the same financing structure on either rail.
Treat the specifics as claims, not as findings. The structure is the part worth understanding, because it turns the network’s own settlement data into an underwriting input and removes the prefunded account that made card programs expensive to run.
What breaks, and who is holding it when it does
Settling in tokens relocates risk instead of removing it, and the new risks land in places card teams are not used to watching.
The first is the interval between authorization and settlement. A card transaction is authorized in dollars and settled later in a token. If the token trades below par in that window, somebody absorbs the difference, and the network’s rules settle that question, not the program’s assumptions. Programs settling in a coin with a thin redemption path should price that gap instead of treating a dollar token as a dollar.
The second is the chain. Settlement obligations do not pause because a network is congested or halted. A program that can only settle on one chain has taken a single point of failure into its treasury operations, which is the strongest practical argument for a network with nine of them and for holding the ability to move between at least two.
The third is redemption, and it is the one most often skipped. Holding a token to settle is fine while the program can convert it to bank money on demand at the size it needs. That capacity sits with the issuer of the coin and with whichever bank stands behind the program, and it is worth testing at volume well before the week it is needed.
How to actually choose
For a program deciding where to settle, four questions get you most of the way.
Which token does the treasury already hold, since converting on every settlement cycle costs spread and introduces a counterparty on the day money is owed? Which chain do the program’s other flows run on, because settling on a ninth chain the operations team has never reconciled is a real cost that no announcement prices. How bad is the weekend, since a program with heavy Saturday spend gains more from Mastercard’s windows than from Visa’s longer chain list. And what does the sponsor permit, which overrides all three?
The honest answer for most programs is that the network is already chosen by the sponsor, the BIN, and the existing acquiring relationships. The decision available is narrower and still worth making: which of the supported coins to hold, which chain to settle on, and whether to finance the gap or keep funding it with idle cash.
What to watch
Three developments would redraw this comparison.
Whether Mastercard publishes a settlement figure, since a number would turn a capability comparison into a scale comparison and is the single disclosure that would tell the market most. Whether Visa’s seven-day pilot leaves Singapore, because always-on settlement is Mastercard’s clearest advantage and Visa has announced a route to closing it. And whether receivables financing spreads beyond one network, since a facility underwritten on Visa’s settlement files works only for programs settling with Visa, and the first equivalent built on Mastercard data would remove a real reason to stay.
Underneath all of it sits a quieter shift. Both networks began by letting partners pay them in tokens. They are now deciding which tokens exist on the menu, which chains carry them, and, in Visa’s case, what credit gets extended against the flow, which is a minting and settlement posture more than a payments one. The settlement rail was the opening move.
Read more: What stablecoin card spend adds up to in 2026 • Rain’s $250 million round prices stablecoin card plumbing