Stablecoin Payments Explained: Rails, Costs and Settlement in 2026

What a Stablecoin Payment Actually Is

A stablecoin payment moves a tokenised claim on a fiat currency — in practice almost always the US dollar — across a public blockchain. The payer sends tokens from a wallet, the payee sees them credited within a block or two, and the issuer stands behind the token with reserves and a redemption promise. For how the tokens hold their peg, see our stablecoins explained guide; this piece is about the payment rail.

Four properties make that interesting:

• Settlement in seconds to minutes. Once a transaction is final on-chain, value has moved. No overnight batch, no provisional credit clawed back later.
• Always on. Chains do not observe weekends, bank holidays or 4pm cut-offs. For treasury teams this often beats raw speed.
• Programmable. Circle's CCTP V2 adds Hooks, triggering automated actions on the destination chain immediately after a transfer completes, per Circle's CCTP page.
• Reachable without correspondent banking. The receiver needs a wallet and a route to local currency, not a chain of nostro and vostro accounts.

Where stablecoins do not beat what you already have

Domestic card payments in a mature market. Cards settle reliably, carry dispute rights customers value, and sit in every checkout your buyers know. A wallet flow trades a solved problem for an unsolved one.

Small consumer purchases. Below the price of a coffee, fixed friction — wallet setup, network selection, gas — dwarfs any fee saving.

Corridors with no good off-ramp. The chain hop is fast; the last mile into a local bank account is where days and fees reappear.

Anywhere you want reversibility. On-chain settlement is close to irreversible — a feature for a merchant tired of chargebacks, a bug for a buyer wanting recourse.

The Stack: Five Layers Between Payer and Merchant

1. Issuer. Circle for USDC and EURC, Tether for USDT, Paxos for PYUSD and USDG, Ripple for RLUSD. The issuer sets the reserve model, the redemption promise and the freeze powers.

2. Chain. Where fees and finality are set. Payment traffic clusters on Tron, Solana, Polygon and Ethereum layer 2s such as Base and Arbitrum.

3. Wallet and custody. Self-custody, MPC wallets or a qualified custodian — where your key-management policy lives.

4. On-ramp and off-ramp. Turning bank money into tokens and back. Circle states that Circle Mint serves institutions such as exchanges, wallet providers and banks, is not available to individuals, is free for those who qualify, and requires KYC and sanctions screening — with onboarding running from a day to a week or more.

5. Merchant settlement. Whether the business holds tokens, converts instantly to fiat, or runs a hybrid — and how that lands in the ledger.

The thing worth internalising: the transfer is the easy part. Sending USDC between addresses is commoditised and cheap. The hard parts are fiat in, fiat out, travel-rule and sanctions obligations, and reconciliation. Every serious vendor here sells ramps and compliance wrapped around a transfer.

Cross-Border and B2B: The Strongest Real Use Case

The case is strongest where the incumbent rail is weakest.

Correspondent banking has been contracting for over a decade. BIS CPMI and Swift data showed relationships falling roughly 25 percent between 2011 and 2020, including a 4 percent drop in 2020 alone, even as cross-border payment volume rose 2 percent and value rose 7 percent that year (BIS, December 2021). Fewer correspondents means longer chains, more intermediaries taking a cut, and thinner coverage for small markets.

Cost data says the same. The World Bank's Remittance Prices Worldwide, Issue 54 (Q3 2025 data), put the global average cost of sending USD 200 at 6.36 percent, down from 6.49 percent in Q1 2025, with the Global SmaRT average — reflecting cheaper available services — at 3.29 percent (Remittance Prices Worldwide). The UN Sustainable Development Goal target is 3 percent.

The Financial Stability Board's G20 targets aim for 75 percent of cross-border wholesale, retail and remittance payments to be credited within one hour by end-2027, with retail global average cost no more than 1 percent and no corridor above 3 percent. In October 2025 the FSB reported its KPIs had improved only slightly since 2023 and that meeting the timetable looked unlikely (FSB).

That gap is the opening stablecoins are walking through. For B2B supplier payments, contractor payouts and inter-entity treasury movement, the proposition is concrete: money that arrives on a Saturday, where the alternative is three business days and a stack of lifting fees.

Tron became a payments rail through cheapness and liquidity. A TRON DAO press release in July 2026, citing Token Terminal data, put USDT supply on Tron above USD 90 billion and year-to-date USDT transfer volume at roughly USD 4.2 trillion (via The Block). Treat that as a foundation's marketing claim, not an audited figure — though the directional point, that USDT on Tron is the workhorse of informal dollar transfer across parts of Latin America, Africa and Southeast Asia, is not contested.

Payment Rails Compared

Two honest reads: inside a modern instant-payment scheme stablecoins usually lose, and their cost advantage lives in the ramps, not the network fee.

The Rails and Infrastructure Worth Knowing

Circle: USDC, CCTP and Circle Mint

Circle publishes reserve holdings weekly plus a monthly attestation under AICPA standards, and holds most USDC reserves in the Circle Reserve Fund, an SEC-registered 2a-7 government money market fund managed by BlackRock (Circle transparency).

For payments, CCTP matters more. It burns USDC on the source chain, waits for a Circle attestation, and mints native USDC on the destination — no wrapped assets, no bridge liquidity pool. Circle's blog records that CCTP V2 became canonical on 14 November 2025, live on 17 chains then, with V1 phase-out beginning 31 July 2026 (Circle); the product page listed 23-plus chains in July 2026. V2 offers a Standard Transfer matching source-chain finality at no protocol fee, and a Fast Transfer settling faster than finality for an on-chain fee.

Tether: USDT and Tron

USDT is the liquidity that actually exists in most emerging-market corridors. Tether publishes circulation data and quarterly attestations prepared by BDO Italia via its transparency page. Note the distinction: an attestation is a point-in-time assertion, not a full audit, and Tether has not published one from a Big Four firm. For an issuer-by-issuer comparison, see our state of stablecoins breakdown.

Stripe and Bridge

Stripe acquired Bridge and now sells stablecoin acceptance as an ordinary payment method. Per Stripe's documentation, it accepts USDC on Tempo, Ethereum, Solana, Polygon and Base, plus USDP and USDG for US businesses; payments settle into the Stripe balance in the merchant's local currency, with a 10,000 USD per-transaction limit; availability is US-general with private preview in the EU, Hong Kong, Mexico and Switzerland. Bridge launched Open Issuance on 30 September 2025, letting a business stand up its own stablecoin with reserves allocated through partners including BlackRock and Fidelity (Stripe); in February 2026 the OCC gave preliminary conditional approval for Bridge National Trust Bank (OCC Decision 1365).

PayPal PYUSD

PYUSD is issued and custodied by Paxos Trust Company and distributed by PayPal, backed by US dollar deposits, Treasuries and cash equivalents, redeemable 1:1, with monthly reserve reports and third-party attestations (PayPal). PayPal also advertises a variable rewards rate — 4 percent at the time of writing — for holding PYUSD in its app. Hold that thought; it collides with the regulation section below.

Visa and Mastercard settlement

On 16 December 2025 Visa announced USDC settlement in the United States, with Cross River Bank and Lead Bank settling in USDC over Solana, and stated that as of 30 November 2025 its monthly stablecoin settlement volume had passed a USD 3.5 billion annualised run rate. Mastercard announced expanded settlement support for regulated stablecoins including USDC, PYUSD, USDG and RLUSD on 3 June 2026, per CoinDesk — we could not retrieve Mastercard's own release, so treat that as press-reported.

Note what these are not. They are the networks settling with their own issuing and acquiring banks. Your checkout is unchanged; the interbank leg stops waiting for banking hours.

What It Actually Costs

Fee comparisons here are usually rigged. Here is a like-for-like one from a single vendor's price list.

Stripe's pricing lists 2.9 percent plus 30 cents per successful domestic card charge, with 1.5 percent added for international cards and 1 percent for currency conversion — against 1.5 percent for stablecoin payments, which Stripe says covers conversion to fiat, wallet and AML screening, fraud prevention and gas sponsorship. Same processor, same merchant, roughly half the take rate.

On the network side:

• Solana charges a base fee of 5,000 lamports per signature plus an optional prioritisation fee derived from compute units, per the Solana docs. At any plausible SOL price that is a small fraction of a cent.
• Tron uses a bandwidth and energy model. Accounts get 600 free bandwidth units daily; energy comes from staking TRX or burning TRX at the energy unit price, which the official docs list at 100 sun per energy unit. The dollar cost of a USDT transfer therefore moves with the TRX price and whether you have staked — historically anywhere from effectively zero to several dollars.
• Ethereum layer 2s such as Base and Arbitrum got much cheaper once blob transactions arrived with EIP-4844, and simple transfers typically cost cents or fractions of a cent — but this is congestion-dependent, so check a live tracker. Our gas fees explainer covers the mechanics. Ethereum mainnet remains the wrong chain for payments on cost grounds.

On the incumbent side, US debit interchange for covered issuers is capped at 21 cents plus 5 basis points of transaction value under 12 CFR 235.3(b). Credit interchange is not capped in the US and runs higher. Wire fees vary so widely by bank and corridor that any single figure would mislead — the headline sender fee is only part of the cost, with lifting fees and the FX spread often exceeding it.

Regulation: GENIUS in the US, MiCA in the EU

United States. The GENIUS Act was enacted on 18 July 2025 as Public Law 119-27. Per the statutory text, a permitted payment stablecoin issuer must have:

• Reserves on at least a 1-to-1 basis in defined eligible assets — US currency, Federal Reserve balances, insured demand deposits, short-dated Treasuries, Treasury-backed repo and government money market funds (Section 4(a)(1)(A)).
• A published redemption policy with clear procedures and fee disclosure, and seven days' notice of changes.
• Monthly publication of reserve composition, examined by a registered public accounting firm.
• No interest or yield to holders. Section 4(a)(11) states that no permitted or foreign payment stablecoin issuer shall pay the holder any form of interest or yield, in cash, tokens or other consideration, solely in connection with holding, using or retaining the stablecoin.

Implementation is live but unfinished. The OCC issued its proposed rule on 25 February 2026, adding a new 12 CFR Part 15, and states the effective date is the earlier of 18 months after enactment or 120 days after final rules from the primary federal regulators (OCC Bulletin 2026-3). The FDIC approved its own proposal on 7 April 2026 with a 60-day comment window (FDIC), and Treasury has proposed illicit-finance rules. Final rules were not all in place when this was written.

The interest prohibition has the sharpest commercial teeth, and it is drafted at the issuer. That leaves an open question about distributors: PayPal advertises a rewards rate on PYUSD balances in its app while Paxos is the issuer. Whether such arrangements survive final rulemaking is unsettled — the battleground covered in our piece on yield-bearing stablecoins.

European Union. MiCA classifies single-currency stablecoins as e-money tokens. Article 50 is blunt: issuers shall not grant interest on e-money tokens, service providers shall not grant interest when providing related services, and any remuneration tied to holding period counts as interest (ESMA single rulebook). EMT rules applied from 30 June 2024 and the rest of MiCA from 30 December 2024. ESMA notes the grandfathering window for service providers under Article 143(3) ran to 1 July 2026 — now behind us. Our cross-border compliance guide maps how MiCA, SEC, VARA and MAS expectations differ.

Risks You Are Actually Taking

Depeg. Reserve backing is a promise about assets, and assets sit somewhere. In March 2023 Circle said that USD 3.3 billion of USDC reserves — roughly 8 percent — were stuck at the failed Silicon Valley Bank, and USDC traded down to around 87 cents before recovering (CNBC, 11 March 2023). A well-run issuer is not a risk-free one.

Freeze and blacklist power. Major fiat-backed stablecoins carry an administrative freeze function in the token contract. Circle's USDC terms reserve the right to block addresses; issuers act on law-enforcement and court orders. Your funds can be immobilised without your consent.

Finality and reorg risk. Chains reach finality differently and at different speeds. Crediting a customer on first confirmation is a policy decision, not a technical given — set a threshold per chain and value band.

Wrong-chain and wrong-address sends. A classic operational loss. The same stablecoin exists natively on many chains; sending to a deposit address on the wrong network is often unrecoverable. Enforce allowlists, chain-aware validation and a small test transaction first.

Compliance and AML. Travel-rule obligations, sanctions screening on counterparty addresses, and licensing in every jurisdiction you serve. This is where projects die.

FX for non-USD businesses. Bill in euros, pounds or rupees while settling in a USD stablecoin and you have taken on currency risk you did not previously have.

Counterparty risk. Ramps fail, custodians have outages, bridges get exploited. Design for a fallback rail.

Nothing here is financial, legal or tax advice. Holding or transacting in stablecoins involves risk, including total loss. Take professional advice for your jurisdiction.

When Stablecoin Payments Make Sense — and When They Do Not

Strong case, run the pilot:

• Cross-border B2B payments where your bank quotes multiple days and unpredictable fees.
• Paying contractors or sellers across many countries via a patchwork of local rails.
• Inter-entity treasury movement across time zones, where weekend availability cuts buffer capital.
• Flows where the counterparty already holds dollars on-chain.
• Interbank settlement, if you are a bank or acquirer.

Weak case, do not force it:

• Domestic checkout in a market with FedNow, SEPA Instant, UPI or Pix.
• Low-value purchases where onboarding friction exceeds the saving.
• Corridors with no licensed off-ramp where the recipient needs local cash.
• Products where customers expect chargeback rights.
• Situations creating USD exposure you cannot manage.

A workable test: if the incumbent rail in your corridor takes more than one business day, costs more than about 1 percent all-in, or does not run at the weekend, stablecoins deserve a pilot. Otherwise they do not.

How This Guide Was Built

This is a desk-research synthesis, not a first-party benchmark. We did not process payments, run transactions or measure fees ourselves; no figure here is our own measurement.

Every number came from a named source, dated in the text and linked above: the GENIUS Act on govinfo; OCC and FDIC rulemaking documents; ESMA's single rulebook for MiCA Article 50; Circle's product, transparency and blog pages; Stripe's pricing page and developer docs; PayPal's PYUSD page; Solana and TRON developer docs; 12 CFR 235.3; the World Bank's Remittance Prices Worldwide; the FSB's G20 targets and October 2025 progress statement; and BIS CPMI data. The Mastercard item is press-reported because we could not retrieve the primary release.

Where a figure moves constantly — gas costs, Tron fees, layer-2 fees, stablecoin supplies — we describe the mechanism and point at the live source instead of freezing a number that will be wrong next week. Where a claim comes from an interested party, such as the TRON DAO volumes, we say so.

Conclusion

Stablecoins are not replacing card networks or domestic instant payments, and anyone selling that story is selling something. What they are doing is eating the worst part of the payment system: slow, opaque, expensive cross-border settlement on a shrinking correspondent-banking network the FSB's own KPIs say is off track for its 2027 targets.

The infrastructure has matured accordingly. CCTP V2 is canonical and multi-chain, Stripe sells stablecoin acceptance at roughly half its card take rate, Visa settles in USDC with US banks, and the GENIUS Act and MiCA have given the asset class a rulebook. The problems that remain are not cryptographic. They are ramps, licences, reconciliation and the last mile into a bank account.

Pick the corridor where the incumbent rail is genuinely bad, pilot there, and measure all-in cost including FX and off-ramp, not the gas fee.

This is an editorial synthesis of documentation, public data, and community reports; see our [methodology](/methodology). Verify current details with each project.