Visa, Stripe, PayPal, Mastercard: Why They Actually Use Solana

Visa, Stripe, PayPal, Mastercard: Why They Actually Use Solana

Solana’s Payment Infrastructure: A Settlement Pipe With Card Networks Bolted On

DividendChase LTD | Institutional Research
As of 29 September 2026

Solana’s payments story is not Solana Pay QR codes in a coffee shop. It is stablecoin settlement behind Visa, Stripe, PayPal, Mastercard, Western Union, Worldpay, and a handful of U.S. banks, plus a new layer aimed at AI agents. The chain is the cheap, always-on hop. The licensed dollar and the card scheme still sit off-chain. Do not price SOL as if it replaced VisaNet.

What the chain actually supplies

Property Why payments teams use it
~400 ms slots, sub-second “confirmed” Weekend and holiday settlement that ACH/wires cannot do
Fees in fractions of a cent High-frequency treasury and payouts
Token Extensions (transfer hooks, confidential balances, required memo) Compliance features PYUSD-class issuers need
Single global state No L2 bridge in the merchant path
Firedancer / dual-client path Outage memory is the historical objection; not fully retired


Throughput share of stablecoin transfers (one Q1 print near three-quarters of count) is velocity. Ethereum and Tron still hold most of the float. Solana holds the hop. That distinction is the whole infrastructure thesis.

The stack, layer by layer

1. Dollar inventory (issuance)
Payments need a unit that is not SOL.

  • USDC (Circle) — wholesale default for Visa settlement and most enterprise APIs.
  • USDT (Tether) — now on bank rails too (Column lists USDC and USDT with Solana as the documented default).
  • PYUSD (PayPal) — Token Extensions; Solana treated as a primary/default network; hundreds of millions of dollars on-chain at various 2025–26 prints (one Feb 2025 figure was ~$835 million on Solana).
  • USDPT (Western Union) — launched on Solana (May 2026 vintage), meant to hit WU’s agent network instead of pre-funded nostro accounts.
  • USDG (Paxos / Global Dollar Network) — Worldpay-linked; a large share of USDG issuance has been cited as Solana-native.
  • Others in the catalog: USDP, FIUSD (Fiserv), tokenized T-bills (BUIDL, BENJI) as treasury, not POS tender.

Issuer law is GENIUS / MiCA / EMI — not Solana consensus. The chain can freeze only what the token program and issuer hooks allow.

2. Open protocols (the “Solana Pay” layer)
Solana Pay (2022) is an open request spec: merchant invoice → wallet signs a USDC transfer. Shopify app, Helio-class plugins, KSNet trials in Korea. Useful, small relative to Visa settlement. It is the public protocol card networks do not need.

Pay.sh (Foundation + Google Cloud, May 2026) is the agent fork: an AI client with a Solana wallet pays per API call (BigQuery, Vertex, Gemini) in stablecoins, no API-key subscription. Standards named in the launch copy include x402 and machine-payment protocols. This is a new demand class, not a replacement for checkout. Mastercard’s Agent Pay for Machines (AP4M) put the Foundation in a 30-partner launch set for the same idea on mixed card + stablecoin rails.

3. Orchestration (the real product companies buy)
Enterprises do not speak Solana Pay. They speak Stripe, Bridge, Crossmint, BVNK, MoonPay, Sphere, Coinflow, Modern Treasury.

  • Stripe added solana as a crypto network in API version 2025-10-29. Checkout can take USDC (and other coins) on Solana; merchant receives fiat in the Stripe balance. Bridge (Stripe-owned) is the stablecoin factory and wallet layer. Tempo is Stripe’s own L1 for new merchant defaults — Solana remains a live rail for balances that already sit there. Meta creator payouts (Colombia, Philippines, April 2026) run USDC on Solana and Polygon through Stripe.
  • Crossmint — enterprise wallets; plumbing behind WU USDPT and Tala-style lending disbursements.
  • Solana Developer Platform (SDP), March 2026 — Foundation-bundled API over 20+ vendors: Helius/Alchemy/QuickNode, Fireblocks/Anchorage/BitGo/Privy, Chainalysis/TRM, Bridge/MoonPay. Early names: Mastercard, Western Union, Worldpay. Modules: issuance + payments now; trading/FX later in 2026. This is how a bank avoids assembling the stack itself.

4. Card networks (wholesale, not consumer “pay with SOL”)

  • Visa: U.S. USDC settlement on Solana from 16 December 2025 with Cross River and Lead Bank. Banks push USDC to Visa’s Solana address instead of a five-day bank cycle. Later prints put Visa’s multi-chain stablecoin settlement at a multi-billion annualized run rate; Solana was the first U.S. public-chain rail, not the only one. Stablecard (WU + Rain, Aug 2026): USDPT-backed Visa in 37 markets — consumer spend, wholesale still VisaNet.
  • Mastercard: Solana among eight chains for 24/7 stablecoin settlement; SDP user; AP4M partner.
  • Worldpay: validator + USDG merchant settlement; claimed ~50% faster processing vs legacy.
  • Column N.A.: FDIC bank; native USDC/USDT APIs with Solana as the documented default; card auth can hit the live stablecoin balance with no float window. SoFi is in the same “chartered bank + Solana volume” set.

5. Payouts and remittances
Gusto + Zero Hash: USDC contractor payouts. WU / MoneyGram: agent cash-out as the off-ramp that makes a Solana dollar spendable in places wallets are not. Huma (post-Arf): on-chain trade credit / SWIFT-replacement volume in the billions of dollars of 2025 flow — credit infrastructure, not POS.

6. Wallets and custody
Phantom / Solana Mobile for retail. Fireblocks, Anchorage, BitGo, Privy, Turnkey, Crossmint for the payment company. Without those names, SDP is a slide.

A dollar’s path (the infrastructure that matters)

Merchant checkout (Stripe-style)
Buyer USDC on Solana → Stripe/Bridge address → optional instant FX → merchant fiat balance. Solana is invisible to the buyer if the UX is good.

Visa issuer/acquirer settlement
Bank obligation in USDC → Solana transfer to Visa → books update same day, including Sunday. The card swipe still hits VisaNet. Only the net settlement moved.

WU remittance
USDPT minted / transferred on Solana → Digital Asset Network → cash at an agent. Working-capital claim: less pre-funded inventory in corridors.

Agent / API
Model calls Pay.sh or AP4M → stablecoin micropayment → Google Cloud or a machine vendor is paid. This is the 2026 product the Foundation is hiring payments GMs to sell.

What is still missing

  • Chargebacks and disputes. Stablecoin transfers are final. Card networks keep the consumer-protection wrapper; on-chain Pay is “push only.”
  • Identity. Token Extensions help; they do not replace KYC at the issuer or CASP.
  • FX inventory. On-chain dollars still meet local fiat at an off-ramp that holds inventory (Stripe, WU, Column).
  • Tempo / other L1s. Stripe can route new volume off Solana. Multi-chain is the processors’ strategy, not monogamy.
  • Finality vs operations. Until Alpenglow certificates are live and Firedancer votes, large treasuries will still add minutes of operational delay on top of slot time.
  • Issuer concentration. If Circle or PayPal paused Solana deposits, the rail would look empty even if TPS was fine.

Investor map (payments sleeve only)

Exposure What you actually own
SOL / BSOL Fee + MEV + narrative on settlement velocity. Not a take-rate on Visa volume.
V / MA Networks that added an extra settlement asset. Tiny Solana beta.
PYPL PYUSD distribution; still a PayPal P&L story.
WU USDPT + agent network option. Execution and regulation dominate.
COIN Exchange + USDC services + listings. Broad crypto.
Circle (if/when listed) The dollar inventory most of the stack settles.
Stripe (private) Closest “orchestration” equity — and the firm building a competing L1.


Protocol tokens (JUP, etc.) are not payment infrastructure. Helius / Crossmint / Bridge are the picks-and-shovels; most are private.

DividendChase stance

Solana payment infrastructure in September 2026 is a three-layer machine: Token Extensions dollars, an open request protocol almost nobody at a bank uses directly, and processors + card schemes that hide the chain. Visa’s USDC settlement, Stripe’s solana enum, PYUSD, USDPT, Column’s bank API, SDP, and Pay.sh/AP4M are real. They make SOL a meter on cheap settlement, not a payments monopoly.

Own the token if you want high-beta throughput. Own V/MA/PYPL/WU if you already own global payments. Do not add them together and call it one Solana position. The pipe can be swapped. The licence cannot.

Intelligence for the Discerning Investor
DividendChase LTD