Arc Goes Live Sept. 16: Circle Wants USDC to Be Both the Money and the Gas
Circle's layer-1, Arc, hits public mainnet on Sept. 16 with a design where USDC pays the network fee from the same balance you're trying to spend. It's a genuinely smart piece of plumbing, but it doesn't fix the $111 billion gap with Tether. Here's what actually matters.
I've got a wallet with about eleven dollars of USDC in it and something like forty cents of ETH, and I can't move the USDC anywhere because I can't cover the gas. That's not a rounding error. That's the entire user experience of stablecoins in 2026, and it's the exact problem Circle is trying to kill on Sept. 16.
Arc, Circle's own layer-1, goes to public mainnet that day. The public testnet opened Oct. 28, 2025, and the private version had more than 100 builders on it by August. Fine, but the launch date isn't the story. The fee design is.
One balance, one asset, no second token to buy
Arc is an Ethereum-compatible layer-1 where USDC pays the network fee. Read that again slowly, because it sounds small and it isn't. On most EVM chains, you need the native token to move anything, which means a merchant accepting dollars has to also hold a volatile asset it never wanted. Arc collapses both into one line item. You hold USDC, you send USDC, and the fee comes out of the same balance. Developers keep their familiar tooling. The documentation promises finality in under a second.
Circle's wallet guidance makes a sharper point than most people caught. The native USDC interface and the token interface on Arc represent the same holding, not two different pots. Any wallet that renders them as separate balances is double-counting the customer's money. That's a detail you only write down when you've watched someone get it wrong in production.
In traditional markets, this would be called vertical integration, and it's the difference between being a settlement layer and being the thing settled in. Visa doesn't just move dollars. It's the reason the rails exist, and it collects on every transaction that touches them. Circle is trying to own both halves of that equation for stablecoins.
Here's the part that deserves more attention than the fee mechanic. Arc pairs open developer access with a permissioned validator set. Anyone can build. Not anyone can validate. The founding cohort includes BlackRock, DTCC, Visa, Mastercard, Standard Chartered, and Circle itself, which is a who's who of the institutions that stablecoins were supposed to disintermediate.
So what you've got is a consortium chain wearing an EVM costume. That's not a criticism. It's a description, and it's probably the right call if your customers are clearing houses and card networks. But the crypto-native crowd should stop pretending this is permissionless in any meaningful sense. Access to build isn't access to govern.
The market math that matters
Circle reported $73.3 billion of USDC in circulation at the end of June. Tether reported roughly $184.6 billion of USDT issued at the same quarter-end. That's a gap of about $111 billion, and no amount of clever fee abstraction closes it overnight.
And this is where I'd push back on the bullish reading. Extra activity on Arc and extra demand for USDC are two different outcomes, and the difference is the whole ballgame. If you bridge your USDC from Base to Arc and pay fees with it, you've moved a balance and established a use case. You haven't grown the float by a single dollar. Circle gets a better story about utility without actually gaining share.
The comparable in TradFi is a bank that figures out how to make its own checking accounts the settlement asset for an interbank network. Useful. Lucrative. Completely unrelated to whether depositors bring more money in the door.
What would actually move the number is boring. Merchants holding working capital in USDC on Arc instead of sweeping to a treasury account every night. Institutions parking settlement balances there because the friction is low enough that the treasury desk stops fighting it. That's a behavior change, not a feature launch, and it takes quarters to show up in circulation data.
One more thing worth flagging. Arc advertises opt-in privacy, and the execution docs still list Arc Privacy Sector and Stablecoin Services as planned and unavailable. Read the documentation, not the announcement. That's true of every launch in this industry and it's true here.
What I'd actually do with this
Two takes, both slightly uncomfortable.
First, the crypto crowd should stop grading Arc on decentralization. It's not that kind of network. It's a regulated settlement utility with a public developer surface and a curated validator list, and it's aimed squarely at the part of the market that will never touch a permissionless chain. Judging it by the wrong scorecard just produces bad commentary.
Second, and this is the one I'd bet on, the fee abstraction is the most underrated product decision in stablecoins this year. Not because it's flashy. Because it removes the single dumbest reason payments fail. The Sharpe ratio on that change is genuinely favorable, meaning the risk-adjusted payoff of fixing basic plumbing beats almost anything else Circle could ship in the same window.
Crypto is pricing in what equities haven't here. Circle trades on the idea that USDC becomes the operating balance of onchain finance. Arc is the infrastructure for that thesis. It isn't the proof of it.
So watch the number, not the headline. If Arc is working, USDC circulation in the fourth quarter lands meaningfully above $73.3 billion and the mix shifts toward institutions holding balances rather than just routing through them. If it isn't, the float stays flat, the volume shows up anyway, and we all learn that fee abstraction is nice and distribution is everything.
Tether's lead was never about technology. It's about where the dollars already sit.