USDC and USDT are both tokens that represent one US dollar, issued by different companies under different disclosure practices, and they are treated as interchangeable far more often than an operator can afford. For holding a treasury balance the choice is mostly about the publisher. For moving money to somebody who has to accept it, the choice is mostly about what that person's bank, exchange or partner will take, which is a different question with a different answer in every corridor.
Both details matter, and they pull in opposite directions often enough that the right policy is usually to support both rather than to pick a side.
Two dollar tokens, two publishers
USDC is issued by Circle. USDT is issued by Tether. Each is redeemable with its issuer at one dollar by parties with a direct relationship, and each trades at or very near a dollar on the open market because that redemption exists. We cover each one on its own in what USDC is and what USDT is, and the vocabulary around them in the stablecoin glossary.
The difference that shows up in a diligence questionnaire is disclosure. The difference that shows up in an operation is where each token is liquid and who is willing to receive it. These are not the same axis, and confusing them is how teams end up with a policy that reads well and fails at the counterparty.
What each publisher actually publishes
Read the two transparency pages side by side rather than taking either company's summary of the other.
Circle states on its transparency page that USDC reserve holdings are disclosed weekly along with the associated mint and burn flows, that a Big Four accounting firm provides monthly third-party assurance that the value of the reserves exceeds the USDC in circulation, and that those reports follow attestation standards set by the American Institute of Certified Public Accountants. The same page names Deloitte and Touche LLP as Circle's independent auditor for the company's financial statements since fiscal 2022.
Tether states on its transparency page that its tokens are pegged one to one with a matching fiat currency and are backed by its reserves, that information about tokens in circulation is typically published daily, and it carries a reports and reserves section alongside the circulation figures.
Those are two different cadences and two different kinds of statement, and a compliance team that has to write a token policy should read both pages on the day it writes it rather than rely on a summary from a vendor, including this one.
The network decides more than the token does
A stablecoin is a pairing, never a single choice. USDT on Tron and USDT on Ethereum are the same token to an accountant and two different assets to an operation: different addresses, different fees, different confirmation behaviour and different sets of counterparties who will receive them.
Which stablecoin runs on which network, in Lumx coverage
Blockchain | USDC | USDT |
Ethereum | Supported | Supported |
Polygon | Supported | Supported |
Base | Supported | Not available |
Tron | Not available | Supported |
Stellar | Supported, on-ramp only | Not available |
Read that table as a constraint rather than as a preference. A counterparty that settles only on Tron is asking for USDT, because Tron is where USDT lives and USDC does not. A flow built on Base is asking for USDC for the mirror-image reason. The token follows the network in both cases, and USDT on Tron covers what that specific pairing means in practice.
In Latin America the counterparty usually decides
The pattern across the region is that USDT carries the settlement volume between businesses and USDC carries more of the institutional and treasury volume, with local exchanges, brokers and OTC desks quoting deeper books in USDT on Tron than in anything else. That is not an argument about quality. It is an observation about where the liquidity sits, and liquidity is what determines whether the rate you quote your customer survives the trade.
This has a direct consequence for product design. If your payout depends on converting a token into local currency at a competitive rate, you want the token the local market actually trades. If your treasury holds a balance for weeks, you are choosing a publisher rather than a market, and the disclosure question outranks the spread.
I do have a preference between the two, and the counterparty overrules it roughly as often as it agrees with me, which is the right way round. Choosing the token for a payout is not a way of announcing which issuer I trust. It is a property of the person receiving the money: they settle in whatever their own bank, exchange or partner will take, and if I insist on my preference against that, I do not win the argument. I lose the corridor. Where I do hold a line is on treating the token and the network it runs on as one decision rather than two. A policy that says "we use USDC" and stops there has not decided anything, because those same four words describe a flow on Base and a flow on Stellar, and once you have to operate them the two have almost nothing in common. So the rule I actually apply has two halves: the treasury decides which issuer to hold, and the corridor decides which token on which network goes out.
Cost is a network question wearing a token costume
The fee to move a stablecoin is charged by the blockchain, not by the issuer, so it varies by network and by congestion rather than by whether the token says USDC or USDT on it. A transfer on Ethereum during a busy period can cost multiples of the same transfer on Polygon, Base or Tron, and for payout batches with many small payments that difference stops being a rounding item.
The second cost is the spread on the conversion into local currency, and that one does follow the token, because it follows the depth of the local market. Converting into Brazilian reais over USDT to BRL and over USDC to BRL are different trades against different books on the same day, and the same holds in Mexico for USDT to MXN.
When the choice does not matter
If the balance never leaves your own custody and never converts, this is a governance question for your board rather than a payments question, and either token answers it. If your volumes are small and your counterparty accepts both, pick the one your provider defaults to and spend the attention somewhere that pays better.
It also does not matter if you have already decided the network for another reason. Building on Base settles the question at USDC. Settling with a partner who only receives on Tron settles it at USDT. In both cases the token was chosen by the earlier decision, and re-opening it is work without an outcome.
Where the choice is wrong is at the extremes. A treasury policy that concentrates a material corporate balance in a single token because it is cheaper to move is optimising the wrong variable. And a payout operation that refuses the token its counterparties settle in is going to discover that principle is expensive.
Holding both and choosing per transaction
Lumx is stablecoin payments infrastructure for businesses that move money between Latin America and the rest of the world: one API to collect, hold, convert, and pay out in BRL, MXN, COP, USD, EUR, and GBP or in USDC and USDT, over local rails such as PIX, SPEI, PSE, ACH, FEDWIRE, SEPA, and Faster Payments, with SWIFT and on-behalf-of payments and collections (POBO and COBO) in USD, EUR, and GBP, plus named virtual accounts, custodial wallets, and KYB/KYC built in.
Every customer gets a wallet on each supported network, and each wallet holds balances per token, so the pairing is an argument on the request rather than a decision you make once at signup. Omit it and the project default applies. Name it and you get the network and the token you asked for, on the same balance, through the same integration.
The recommendation, stated plainly because a comparison that ends in "it depends" is not worth publishing: hold treasury in USDC if your governance framework cares about disclosure cadence, and pay out in whatever pairing your counterparty settles in, which in Brazil and Mexico is frequently USDT on Tron. Where we are not the right pick is when you want one token on one network and nothing else, forever, with no local conversion: that is a custody problem rather than a payments problem, and a wallet provider will serve you better than we will.
Methodology and sources
Reserve and disclosure statements are quoted from the publishers' own transparency pages, both read on September 24, 2026: circle.com/transparency for USDC and tether.to for USDT. Neither figure is restated from a third-party tracker, and no circulation or market-share number appears in this post, because those move faster than a blog post is revised and the comparison does not turn on them.
Network and token availability is from the Lumx coverage matrix, read on the same date, and describes what is supported on this platform rather than what exists in the market. Both tokens run on networks we do not support, and both are held by counterparties we do not touch.
The statement about regional settlement preference is an operator observation from running these corridors, not a measured market share. It is stated as a pattern for that reason, and a team with a specific counterparty should confirm it with that counterparty rather than with us.
Verified on September 25, 2026. Operational context, not legal, tax, or investment advice.
Cover photo: Tiago Gerken on Unsplash.
Is USDC safer than USDT?
They publish different things at different cadences, and the honest answer is that your risk framework has to decide what that is worth. Circle discloses reserve holdings weekly and publishes a monthly third-party assurance report under American Institute of Certified Public Accountants (AICPA) attestation standards. Tether publishes circulation data daily alongside its reports and reserves section. Read both pages directly before writing a token policy.
Which one should we use for payouts in Brazil and Mexico?
Whichever one your receiving counterparty settles in, which in both countries is frequently USDT on Tron. The conversion into local currency happens against the local market, so the token with the deeper local book usually produces the better rate for your customer. If the counterparty accepts both, the decision moves back to your own treasury preference.
Does the blockchain matter as much as the token?
Usually more. The same token on two networks has different fees, different confirmation behaviour and different sets of counterparties willing to receive it, and some pairings simply do not exist. Treat the token and the network as one decision rather than two.
Can we hold both without running two integrations?
Yes, and that is the normal setup. A customer wallet exists on each supported network with balances per token, so the pairing is chosen per transaction rather than per account. Teams that support both typically hold treasury in one and pay out in whatever the corridor demands.
