Most stablecoin transactions look simple on the surface. You send USDC from your wallet to someone else's wallet, the transaction confirms, and the balance updates within a few seconds. What's actually running under that interaction is a chain of attestations, mint-and-burn operations, bridge contracts, and reserve attestations spread across multiple blockchains.
Understanding the stablecoin payment rails behind those transfers matters if you build anything in crypto. It also matters if you want to know why some transfers cost a few cents and others cost a few dollars, why settlement sometimes takes two seconds and sometimes two days, and why Tether and Circle can freeze tokens they issued years ago without anyone's permission.
Two issuers, two very different stacks
USDC and USDT aren't interchangeable products. Circle issues USDC through a regulated entity in the United States. Tether issues USDT through a company that's been settled with regulators in multiple jurisdictions and now operates mostly from El Salvador and other offshore-friendly hubs.
The issuers are different, but the way tokens move is structurally similar. Both follow a mint-and-burn model. When you deposit a dollar with Circle, they mint a fresh USDC token on whichever chain you asked for. When you redeem USDC, they burn the token and send you a real dollar back through a bank wire.
That mint-and-burn design means issuers need to keep enough dollars in reserve to cover every outstanding token. Circle publishes monthly reserve attestations from a Big Four accounting firm. Tether publishes attestations from a smaller Italian firm, and the contents of those attestations have been the subject of regulatory action by the CFTC and the New York Attorney General.
From a developer perspective this matters because not every stablecoin in your wallet is fully reserved. Always check the attestation before you accept payment.
How a single USDC transfer actually flows
Say you send 100 USDC to a friend on Ethereum mainnet. Here's what happens between clicking "send" and your friend seeing the tokens:
- Your wallet signs a transaction calling the USDC smart contract's
transferfunction. - The transaction hits the Ethereum mempool. Validators pick it up, include it in a block, and write the new state to the chain.
- The USDC contract checks your balance, subtracts 100, and adds 100 to your friend's address.
- Roughly 12 seconds later on Ethereum, 400 milliseconds on Solana, or near-instant on Base, your friend's wallet picks up the event and updates the UI.
The smart contract is the entire settlement layer. There's no clearinghouse in the middle. That's what people mean when they say stablecoins settle onchain.
The cost varies dramatically by chain. On Ethereum L1, a USDC transfer might cost $1 to $5 during peak congestion. On a Layer 2 like Base or Arbitrum, the same transfer costs a few cents. On Solana, it's often under a tenth of a cent. This is one reason stablecoin billing on Solana caught on for micropayments and SaaS subscriptions.
Bridges, wrapped tokens, and the multi-chain problem
A USDC on Ethereum isn't the same token as a USDC on Solana or on Avalanche. Each chain has its own contract with its own address and its own ledger. If you want to move the same dollar's worth of USDC from one chain to another, you're not just sending a message, you're either bridging or doing a burn-and-mint swap.
The most common bridge pattern locks your USDC on chain A and mints a wrapped version on chain B. When you bridge back, the wrapped version gets burned and the original gets unlocked. The bridge holds the locked tokens in a custodial wallet, or in a smart contract controlled by a validator set.
This is where the largest stablecoin hacks have happened. The Ronin bridge lost $625 million in 2022 because the validator set was effectively five keys held by the same team. The Wormhole bridge lost $320 million the same year because of a signature verification bug.
Today the safer pattern is canonical USDC. Circle signs a message that the Circle bridge contract uses to mint a real USDC on the destination chain directly, without locking your tokens first. The USDC on Ethereum burns and a fresh USDC appears on Base. No wrapped token, no bridge liquidity pool, no third-party validator set in the middle. That's now how most institutional USDC flows between chains.
Tether has historically been slower here. USDT on most chains is still a wrapped version or a separate issuance, and the official cross-chain story is weaker.
Issuer-level controls you should not ignore
Both Circle and Tether can blacklist addresses. When they do, the affected wallet can still hold the tokens, but it cannot transfer them. The contract refuses any transfer that touches a blacklisted address as sender or receiver.
This isn't theoretical. In 2022 Circle blacklisted over $150,000 of USDC held by Tornado Cash users. Tether has blacklisted thousands of addresses associated with hacks and sanctions lists. If you accept USDC as payment from a counterparty, you're trusting that neither you nor they are on a future blacklist.
For DeFi protocols that custody user funds, blacklist resistance matters more than most people admit. Stablecoins on Ethereum that survive sanctions screening while keeping censorship resistance are a smaller subset than you'd think.
Why reserve quality still matters
A stablecoin is only as good as what's behind it. Circle holds USDC reserves in short-dated US Treasuries and cash equivalents. Tether holds a mix of Treasuries, commercial paper, secured loans, and other assets that have evolved over the years.
The mix matters because if reserves are in something illiquid, redemption requests during a crisis can fail. This was the structural risk that hit UST in 2022. It's also why USDC briefly lost its peg in March 2023 when Circle had $3.3 billion stuck at Silicon Valley Bank. Within 48 hours Circle confirmed the funds were recoverable and USDC re-pegged. That round-trip is the kind of evidence you want to see before treating a stablecoin as cash-equivalent.
Reserve quality is also why MiCA's stablecoin rules in Europe focus on issuers, not users. The framework only allows issuers that meet specific capital, reserve, and audit requirements to operate there.
When stablecoins beat ACH and when they don't
Stablecoins settle 24/7. There's no weekend cutoff, no cutoff time, no batching window. Settlement is final in seconds on most chains instead of one to three business days on ACH.
For cross-border payments, that's a structural win. A USDC transfer from Argentina to the Philippines doesn't go through a correspondent banking network. It goes through the Ethereum or Solana state machine. There's no SWIFT fee, no intermediary bank, no FX margin extracted at multiple layers.
But stablecoins still lose on several fronts:
- Chargebacks: there are none. Once it's onchain, it's onchain.
- Consumer protections: debit cards have Regulation E. Stablecoins do not.
- Accounting and tax treatment: every onchain movement is a taxable event in most jurisdictions, which makes "stablecoin as checking account" an accounting nightmare.
- Onboarding: buying USDC for the first time still requires KYC at a centralized exchange, which defeats the censorship-resistance pitch.
For B2B settlement between companies that already operate in crypto, stablecoins are now the default. For consumer payments, the rails are improving but the surrounding experience still isn't there.
What's actually changing in 2026
A few shifts worth tracking:
- Circle Payments Network, launched in pilot form last year, is now settling between participating institutions using regulated USDC. This is the closest stablecoins have come to replacing correspondent banking.
- Stripe's stablecoin support for merchants went live in late 2025 and now handles hundreds of millions in monthly volume, mostly in USDC on Base.
- Tokenized money market funds from BlackRock and Franklin Templeton are settling against USDC on Ethereum L2s, treating USDC as cash leg and the fund shares as yield-bearing collateral.
- Solana's stablecoin volume has crossed Ethereum's on monthly basis, driven mostly by retail trading and bot-based settlement.
None of this means stablecoins will replace fiat. It does mean that for the first time, a digital dollar is a working rail for actual economic activity, not just crypto speculation. And that shift has more to do with issuers, regulators, and bridge design than with any individual blockchain.
