Banks on Solana: census of 18 commitments and 2 expired deadlines

Of the eighteen institutional commitments to Solana recorded in this census, eleven were settling real value as of August 8, 2026, three were memoranda signed in Korea within a three-month period, and two —R3 and Fiserv— had passed their public deadlines without confirmation. The registry spans sixteen months of announcements from Visa, Meta, SoFi, BlackRock, three Korean payment companies, a Singaporean bank, and an SBI Holdings subsidiary, ranging from the $3.5 billion annualized run rate that Visa reports settling in USDC to the 8.77 million of Western Union's USDPT in circulation, against a base of $15.66 billion in tokenized dollars on the network. This article organizes each commitment into four verifiable statuses —production, pilot or proof of concept, memorandum of understanding, and announced without confirmation within the deadline— featuring the data supporting each category as of August 8, 2026. It also uses the case of Triple-A, the Singaporean payment gateway that was already settling operations on Solana and lost $11.8 million in a drain of its hot wallets on July 24 and 25, to explain why the stretch between signing an agreement and settling real payments is the most expensive to cross.

Who is moving real money on Solana today?

The largest case by volume is Visa. On December 16, 2025, it announced that Cross River Bank and Lead Bank began settling with it in USDC —Circle's stablecoin, a tokenized dollar backed one-to-one— on Solana, with availability seven days a week, including holidays. Visa then estimated its crypto settlement at an annualized rate of 3.5 billion dollars and announced its expansion "throughout 2026" following the approval of the U.S. GENIUS Act. The program positions issuing banks to close their daily positions with the card network in a digital asset.

The most massive case by potential users is Meta. Since April 29, 2026, it has been paying content creators in USDC directly to their wallet, on Solana and on Polygon, with Stripe as the payment infrastructure provider. The perimeter defines the real scale of the program: it is open only to creators from Colombia and the Philippines, and the company itself clarified that it does not issue any stablecoin of its own.

The production list continues with SoFi, which on May 27, 2026, launched SoFiUSD (SOFID) on Ethereum and Solana, presenting it as the first stablecoin issued by a U.S. national bank, accessible from its app to approximately 15 million users. Then there is B2C2, the SBI Holdings subsidiary that in early April 2026 designated Solana as the primary network for institutional stablecoin settlement—USDC, USDT, PYUSD, USDG, USD1, EURC, and FDUSD—with clients such as Robinhood, Standard Chartered, Anchorage Digital, and Bitget. Also, Singapore Gulf Bank and its USDC issuance and redemption service against dollars, operational since April 17, 2026. And on the tokenized funds side: BUIDL, the BlackRock money market fund with Securitize, arrived on Solana on March 25, 2025—the oldest commitment in this entire census—and maintains $675.0 million on the network, 25.1% of the total fund; Franklin Templeton's BENJI has covered eight public chains including Solana since Q1 2026, while its tokenized assets grew from $594 million to over $2.5 billion.

The native issuer with the steepest curve is USDGO, from Anchorage Digital and OSL Group: it started on Solana in February 2026 with $50 million and by August 8 reached $1.14 billion, nearly 23 times the initial figure in six months and 100% of its circulation on this network.

What does the census look like and what does each status mean?

The four states in the table are defined as follows, with each commitment occupying a single cell based on the evidence available as of the cutoff date; nuances —closed perimeter, security incident, pending tranche— are noted alongside the state where applicable:

• Production. Value is being settled on the network, verifiable on an explorer or a circulation aggregator, or confirmed by the counterparty assuming the risk. When volume is marginal compared to the announcement, it is marked as residual production.
• Pilot or Proof of Concept. The code works, but the scope is closed: testnet, a handful of merchants, a single geographic corridor. There is no commitment to scale.
• Memorandum of Understanding. A document declaring the intent to work together which, barring specific clauses, does not obligate either party to deploy anything.
• Announced without confirmation within deadline. A public date was set, the date passed, and there is no public evidence of compliance as of August 8, 2026. This category does not judge intent; it only records the calendar.

Why is it so hard to move from MOU to production?

Signing an MOU costs a press release. Settling payments costs building an operation that can withstand an audit, and that is where three problems arise that cannot be solved with a marketing budget.

The first is hot custody and its schedule. To pay in real-time, you need a wallet connected to the internet that signs transactions without human intervention—a hot wallet—and that operational balance is available, and exposed, at three in the morning on a Sunday, because a public network settles seven days a week. This is not a hypothesis: the Triple-A drain began on Friday, July 24, 2026, and lasted for 31 consecutive hours, with deposit accounts open while it occurred. Sizing that balance, and the monitoring shifts it requires, is a risk decision that no one makes during the MOU phase.

The second is key management. A regulated entity cannot have a private key on a server: it needs hardware security modules, quorum signature schemes—multiple people or systems authorizing each operation—periodic rotation, and a documented procedure for the day an employee with access leaves. All of this is new infrastructure that does not exist in the traditional banking core.

The third is regulatory compliance. Every transfer to an external address requires checking sanctions, applying travel rules for the originator and beneficiary, and maintaining accounting traceability in a format the auditor understands. The U.S. GENIUS Act—the federal stablecoin law passed in 2025—and the European MiCA regulation, which only applies in the EU, also set reserve and reporting requirements that force a reconstruction of internal processes, not just an API integration.

What does the Triple-A case teach us about operational risk?

Triple-A is a payment gateway licensed by the Monetary Authority of Singapore that was already operating with Solana: its public documentation lists the network among those compatible—alongside Bitcoin, Ethereum, Tron, Polygon, and Arbitrum—and its integration with GrabPay allows users to top up the service's wallet with crypto assets and stablecoins. It is exactly the profile of a company that the institutional map places in the production column.

On July 24 and 25, 2026, its hot wallets suffered a drain across six chains—Ethereum, Tron, Polygon, Arbitrum, Solana, and TON. Initial estimates from security firm PeckShield spoke of $9.7 million; the final count published by specialized media on July 27 raised the figure to $11.8 million after 31 hours of continuous draining. The company stated that customer funds were not affected because they are held in third-party custody, and that losses would be covered by its own reserves. This article does not attribute the incident to any actor: as of the closing date, there is no confirmed public attribution.

The figure to remember from this episode is 31 hours. A drain that extends for more than a full day describes a gap between the moment funds start leaving and the moment someone cuts the flow, and that gap is the pattern we analyzed in detail when studying the distance between detection and patching in 2026 hacks. For those reading the map, the production column groups entities that already maintain an operational balance connected to the internet twenty-four hours a day, with the cost of continuous surveillance that this imposes.

Which announcements have gone months without confirmation?

Two commitments with public dates have already passed their deadlines without confirmation. On December 12, 2025, R3—the banking blockchain software firm and creator of Corda—announced the launch of the Corda protocol, curated yield vaults on Solana-native real-world assets, on behalf of the R3 Foundation, with over 30,000 pre-registrations according to coverage and a stated deadline of H1 2026. On August 8, 2026, the protocol's own website was still offering only a waitlist and a message that the launch will arrive "in 2026." The second is Fiserv: its CEO stated on June 3, 2026, at an investor event, that FIUSD—the stablecoin the company targets at its base of some 10,000 financial entities, with Paxos and Circle infrastructure—would be operational in July. As of August 8, it does not appear with circulation in the aggregator listing 413 stablecoins, and public tracking returns no confirmation of an effective launch.

None of these checkboxes imply a breach of contract or bad faith: a timeline announced at an investor conference is a forecast. The value of tracking them lies in having the schedule at hand on the day each milestone is resolved.

How many tokenized dollars are actually on Solana?

The aggregate from the DeFiLlama public API for Solana on August 8, 2026, was $15.66 billion in dollar-pegged stablecoins. The ten largest issuers account for $15.579 billion of that figure, and USDC and USDT alone make up 64.8% of the total. The distribution by issuer tells a story that the aggregate hides: what portion of each issuer's business flows through this network.

Four of the listed issuers exist solely on this network: USDGO, Solstice's USX, Western Union's USDPT, and JupUSD—the stablecoin of the ecosystem's largest liquidity aggregator, with $52.6 million. At the other extreme, Tether holds only 1.6% of its $183.1 billion global total here. The combined reading is that Solana concentrates new native issuance while remaining a secondary venue for traditional issuers; we explore the contrast with the ecosystem's native liquidity in our analysis of Jupiter as Solana's superapp.

In tokenized real-world assets, we must step down in data quality, and we state this before providing it: the RWA.xyz dashboard returns a 403 error to any automated query, so we are not cross-referencing against the primary source here. What follows are July 2026 readings published by half a dozen media outlets citing that dashboard and agreeing with each other: an all-time high of $3.41 billion in the middle of the month, a late-month reading of $3.7 billion—behind Ethereum ($15.9 billion) and BNB Chain ($4.0 billion)—and a market share of around 10.4%. This is the only block in this article whose figures we have not been able to verify against their origin. We cover the tokenized stocks component of that aggregate in detail in the xStocks analysis.

Where do MoneyGram, Ripple, and the euro fit into the census?

Four boxes require a marginal note to be placed in the correct column of the ledger.

MoneyGram features in infrastructure, with issuance on another network. On June 22, 2026, it joined the network's validator set and the Solana Developer Platform, citing its 60 million customers. Twenty days earlier, on June 2, it had launched its MGUSD stablecoin on Stellar, with Bridge —Stripe's subsidiary— as the issuance partner. The infrastructure commitment and the issuance decision point, in this case, to different networks.

RLUSD remains excluded from the census. Ripple's stablecoin, with 1,453 million dollars in circulation, distributes its issuance between Ethereum and the XRP Ledger depending on the aggregator consulted. On Solana, there is XRP wrapped by a third-party custodian, an instrument with a different issuer and risk profile.

The euro slot is occupied by EURC. Circle's EURC holds 52.7 million euros on Solana, representing 11.6% of its 454.7 million global supply distributed across ten chains. It is the largest euro circulation on the network, far ahead of Societe Generale-FORGE's EURCV, which, with 639,884 euros, holds 0.4% of its 152.8 million here. The French bank deployed the two tokens it promised; volume has not yet followed the deployment.

The Toss Bank memorandum has a corporate context. The June 19, 2026, agreement to pilot remittances coincides with the IPO preparations of its parent company, Viva Republica, in the United States, as noted by Blockhead on July 1. This observation adds a corporate incentive to the timeline. Regarding the third Korean memorandum, from KSNet on July 30, the mechanism it will pilot is of particular interest: Solana Pay and x402, the payment standard for agents whose actual volume we analyzed in our review of x402 on Solana and Base.

How to verify the next "bank X adopts Solana"?

The pattern of the eighteen cases in the census can be compressed into five checks that any reader can perform in five minutes:

• Search for circulating supply or an address. If the asset exists, a public aggregator or a block explorer will show it. As long as the figure does not appear, the commitment remains in the announced category.
• Ask who is providing custody. The issuer, custodian, and gateway operator are usually three different entities, and the risk is distributed unevenly among them.
• Note the committed date. When an announcement includes a deadline, check back when it expires. The two expired cases on this map are detected with that single discipline.
• Check the perimeter. "Meta pays in stablecoins" and "Meta pays in stablecoins to creators in Colombia and the Philippines" describe the same event with incomparable scales.
• Distinguish issuance from infrastructure. Operating a validator, joining a foundation, or signing a memorandum does not imply moving a single customer dollar through the network.

Between now and autumn, four categories will resolve themselves, and one has just moved. USDPT went from $6.82 million to $8.77 million between August 3 and 8—nearly 29% in five days—yet it remains 0.06% of the network's tokenized dollars: it is growing fast on a tiny base, which is precisely how a new product signifies nothing yet. Three remain. Whether FIUSD appears with public circulation, given the July deadline its CEO provided has passed. Whether the Corda protocol opens beyond a waitlist that on August 8 was still asking for registration for "early access." And whether Singapore Gulf Bank's retail phase moves beyond the announcement, thirty-nine days after that "mid-2026" target. We also discuss this context of regulated banks entering crypto via licensing in the analysis of Coinbase's national trust bank charter. And for those arriving at this map without prior network context, the foundation is in our guide to DeFi on Solana.

Sources and links: DeFiLlama — Public Stablecoin API (Consulted Aug 8, 2026) · DeFiLlama — Aggregate by Chain · CoinDesk — Visa settles in USDC with U.S. banks (Dec 16, 2025) · Cross River — Note on Visa settlement pilot · Decrypt — Meta pays creators in USDC via Stripe (Apr 29, 2026) · Cryptonomist — USDPT launch on Solana (May 4, 2026) · CoinDesk — SoFi launches SoFiUSD (May 27, 2026) · B2C2 — Solana as primary settlement network · CoinDesk — BUIDL arrives on Solana (Mar 25, 2025) · CryptoBriefing — Franklin Templeton tokenized fund growth · CryptoBriefing — USDGO surpasses $1 billion · SG-FORGE — EURCV deployment plan on Solana · SG-FORGE — USDCV on Ethereum and Solana · CryptoTimes — Shinhan Card MOU (Apr 30, 2026) · Blockhead — Toss Bank pre-IPO context (Jul 1, 2026) · Solana Compass — KSNet MOU (Jul 30, 2026) · PR Newswire — MoneyGram joins as validator (Jun 22, 2026) · Fintech News Singapore — GrabPay top-ups via Triple-A · Triple-A — Supported networks (Support documentation) · Bitcoin.com — Initial PeckShield estimates on Triple-A · CryptoTimes — $11.8M losses after 31 hours (Jul 27, 2026) · GlobeNewswire — Corda protocol announcement (Dec 12, 2025) · Corda protocol — Website status on Aug 8, 2026 · Payments Dive — Fiserv places FIUSD in July (Jun 3, 2026) · CryptoBriefing — Solana RWA all-time high