Stablecoin issuance has been the one piece of the market a bank could not buy off the shelf. But since July 16, Visa clients have been able to mint, move and manage a dollar token from inside the vendor relationship their risk committees approved decades ago.
The beta of Visa’s stablecoin platform supports one token, Open USD, on three blockchains: Ethereum, Solana and Tempo. Access requires an existing Visa Access ID and business identification number, which closes the door to the general crypto market, according to launch coverage from Bloomberg and CoinDesk.
For banks, the launch reframes the board-level question. It is no longer whether a stablecoin strategy requires a digital-assets division, protocol engineers and a custody build. It is whether the strategy requires owning anything at all.
What the platform actually does
Underneath the branding, the platform is a managed issuance and treasury stack. A client can mint a token and redeem it, hold balances in wallets the platform operates, move value across the supported chains without running its own nodes or keys, and convert between tokens and fiat through built-in ramps.
The piece no startup can replicate is the last one: the apparatus connects to existing Visa payment flows. The pitch is not that stablecoins are new. It is that a bank should not need a new department to use them.
Visa’s own framing of the sale is blunter than the marketing around it. “Stablecoins are opening up a new layer of programmable money, but for most institutions the hard part isn’t the concept, it’s the operational reality,” said Jack Forestell, Visa’s chief product and strategy officer, in the launch announcement.
“With the Visa Stablecoin Platform, we’re giving our clients a single place to mint, move and manage stablecoin operations with the controls, security and network reach they already expect from Visa,” Forestell said. “It’s how we help them turn interest in stablecoins into real products and real payment flows.”
The timing does a lot of work. The beta arrived weeks after Open Standard, the consortium behind Open USD, went public, and weeks before the Treasury Department opened comments on the rules defining who may issue payment stablecoins in the U.S.
Whether or not that sequencing was choreographed, it functions as choreography. The consortium supplies the asset, the platform supplies the machinery and the rulemaking supplies the license that platform clients will need to use either one.
Three stages in five years
Visa has been building toward issuance since 2021, in three stages. Settlement came first, a pilot accepting USDC from a crypto issuer that widened until token settlement spanned nine chains and both sides of the interchange relationship. Card programs came second through the Bridge partnership, now heading toward 100 countries.
Issuance is the third stage, and the one where the money is born. Each stage made the next cheaper to build: settlement taught the network to hold tokens at obligation scale, and cards taught it to convert them at authorization speed.
The build has a traceable arc for readers keeping score. Visa accepted its first USDC settlement from a crypto issuer in 2021, crossed the pilot to merchant acquirers in 2023 and took U.S. domestic stablecoin settlement live in December 2025. Tempo reached mainnet March 18, Open Standard surfaced in June with more than 140 members and the platform beta opened July 16.
Why the gate is the strategy
The client gate has been read as caution. Segmentation is the better reading. The institutions Visa wants minting money on its platform are its existing base of banks, processors and large fintechs, which are already licensed, already underwritten and already inside the network’s rulebook.
There is a regulatory logic underneath the commercial one. The GENIUS Act’s licensing regime takes effect in January 2027, after which issuing a payment stablecoin in the U.S. requires a federal or state license. A platform whose users are all regulated financial institutions maps cleanly onto that law.
How cleanly a network-operated console fits that perimeter is one of the questions the comment period exists to argue. The definitions Treasury lands on will decide whether platform clients license the activity, the console licenses it for them, or both answer for it together.
The regulatory calendar now runs the clock. The GENIUS Act’s licensing perimeter closes in January 2027, making the platform’s clients exactly the population allowed to mint, and the law’s service-provider phase-in ends in July 2028. The Treasury comment window on the issuance definitions is open now, while the beta onboards.
The consortium token goes first
The choice of first token sets up the competitive story. Open USD is the stablecoin of Open Standard, a consortium of more than 140 businesses assembled around Visa, Mastercard and Stripe, with BlackRock among the founding partners named in launch coverage. Members can mint and redeem the token without fees or volume limits.
Stacked together, the pieces form a full vertical: a consortium token whose economics favor its members, issued through a Visa-operated console, integrated into Visa settlement flows. Launch coverage framed the platform as fresh competition for Circle Internet Group, and the framing is fair.
Circle answered before the question was asked. Its Circle Payments Network grew 76% quarter over quarter to reach $14.7 billion in annualized transaction volume in the second quarter, with 175 institutions enrolled, up 29% from the prior quarter, according to the company’s results. Its own blockchain, Arc, reaches mainnet Sept. 16.
The competitive read does not need a research note to state. Network-operated issuance does not attack USDC’s float; it attacks the assumption that the next institutional token defaults to a Circle mint, and Circle’s counter is to make its network compound faster than Visa’s console can fill.
Three ways for a bank to buy a mint
The issuance layer now offers banks three models: buy it from your network, buy it from a specialist such as Paxos, or buy the outcome from Circle and let the vendor keep the keys. Banks get to choose their dependency. None of them gets to avoid having one.
For the specialist infrastructure firms, the beta is validation and warning in one. Bridge before its acquisition by Stripe, Paxos with issuance-as-a-service and Circle with its own mint all told the market that institutional issuance was the prize. The platform confirms the market exists and claims the distribution advantage in it, because the client relationships the specialists spent years courting are the ones Visa already holds.
Tether competes for none of this. Its roughly $183 billion of supply lives mostly outside the licensed perimeter these platforms serve, according to DefiLlama, and an issuance console for regulated institutions neither threatens that franchise nor touches it.
Nothing in the architecture prevents additional tokens, and a platform that only ever mints its sponsors’ asset will look like a walled garden to the treasurers it courts. Whether USDC, PYUSD or bank-issued tokens appear on the supported list is the single most informative signal the platform will send.
Read more: New payment chains court banks with rails built for dollars • Circle turns USDC into a payments network with rails of its own
What the announcement leaves out
The absences in the announcement matter as much as the contents. No bank has been named as an issuing client. No existing major stablecoin is on the supported list. Mastercard, co-parent of the consortium token, runs its own parallel settlement stack rather than sharing this one.
A platform launch with one token and zero named clients is an architecture announcement, and architecture announcements get graded later, on tenants.
BlackRock’s presence connects the platform to the other half of the tokenization story. The asset manager already runs BUIDL, the largest tokenized money-market fund, and manages reserves for major stablecoin issuers. A consortium token whose reserve management can pass through the world’s largest asset manager arrives with its Treasury operation pre-solved.
It also hints at the platform’s likely reserve architecture. An issuance console is, underneath everything, a reserve-administration business: every minted token is a liability that has to be matched, invested and attested. Visa operating the mint while institutional-grade managers run the collateral is the configuration the GENIUS Act was drafted to license.
The chain list and the Stripe convergence
The chain list carries its own signal. Ethereum and Solana are the obvious picks, the deepest liquidity and the fastest consumer rails among public chains. Tempo is the interesting one: the payments blockchain incubated by Stripe with Paradigm, live since March 18, with Visa among its design partners.
The two companies compete hard at the merchant layer and keep converging at the infrastructure layer. Stripe’s Bridge issues cards on Visa rails, Visa’s platform supports Stripe’s chain and both sit inside the same token consortium. A payments industry whose giants compete on product while co-owning the substrate is, a historian would note, how the card networks themselves started.
The dependency bargain banks know
Banks know the bargain on offer from experience. Core processors made deposit-taking easy and made switching processors nearly impossible, and card processors did the same for issuance. A bank that mints through the network’s console inherits convenience and dependency in the same gesture.
No academic literature is needed to price that trade, because the industry’s own history is the dataset. Convenience won every prior round, and dependency was the fee it charged.
The beta’s design also has a shadow. Midsize banks, credit unions and regional processors hold Visa relationships and could clear the gate, but the operational assumptions favor the top few hundred institutions with treasury teams that can manage token float. Below them sits a long tail whose stablecoin exposure will come through settlement and card products built by others.
The platform concentrates capability where capability already was. That is how network economics usually go, and it is worth saying plainly because the marketing will not.
The unpublished terms
Three commercial questions have no public answer yet, and each will matter more than any feature. Custody and liability come first: whose balance sheet stands behind platform wallets, and how a loss would be allocated among Visa, the client and any subcustodian. Institutions will not mint against a terms-of-service shrug.
Economics come second. Nothing has been published on pricing, and the difference between a platform fee and a share of reserve income is the difference between a service and a partnership. Geography comes third: issuance is now a licensed activity in every major market, so the platform’s country list will trace the map of stablecoin law rather than lead it.
The skeptic’s file has three exhibits. Beta gating means zero public volume, and incumbent-built crypto platforms have died quietly in pilot before. The single-token launch invites the walled-garden reading. And bank issuance decisions move at examination speed, so the gap between console access and a chartered bank actually minting could swallow quarters.
None of that reverses the significance. It prices it. The launch is architectural, architecture compounds slowly and a fair measure of the platform will not be available before mid-2027.
What that fair measure looks like is worth writing down now, before the marketing defines it. Tenant count: how many named institutions issue or settle through the console, and of what size. Token breadth: whether the supported list grows past the sponsors’ asset. Disclosed volume: minted supply or settlement flow reported in a form an analyst can track quarter over quarter. A platform that clears all three by mid-2027 changed the market. One that clears none was a press release with a console attached.
Three words coverage will blur
Coverage of the platform will blur three things that are not the same. Issuance means minting and burning tokens against reserves, the legally weighty act the GENIUS Act licenses. Orchestration means moving value across chains and between token and fiat, where most of the engineering lives. Settlement integration means using platform-held tokens to meet Visa network obligations, which the beta gestures at but has not confirmed at scale.
Progress on each will be reported as progress on all three. They are not the same thing, and the platform’s real maturity will show in which of the three carries disclosed volume first.
What happens next
Three markers, in likely order. Additional tokens on the supported list, which would show the platform is infrastructure rather than an Open USD delivery vehicle. A first named bank issuing through it. And general availability beyond the Access ID gate, which would tell the mid-tier institutions currently outside the operational assumptions when their turn arrives.
The deeper question the beta leaves open is what Visa’s own settlement stack does with platform-issued tokens. The network already pilots stablecoin settlement across nine chains, and tokens minted on the platform settling network obligations natively would close the loop. Every incentive in the design points that way, and no one at the company has yet said whether it will.