Arc Is Live: The Design Choices Behind Circle's Dollar-Native Chain

Circle switched on the public mainnet of Arc today. Arc is a Layer-1 blockchain built by the company that issues USDC, and its 11 founding validators include BlackRock, Visa, Mastercard and DTCC.

This post goes through the main design choices behind it, what each one gives up, and what the launch means depending on who you are. First, a few terms.

A few terms

Layer-1 (L1). A base blockchain with its own validators and consensus rules, like Ethereum or Solana. An L2 is a separate network built on top of an L1 to make transactions faster and cheaper.

Gas token. The asset you pay transaction fees in: ETH on Ethereum, SOL on Solana. Chains charge fees because without a cost per transaction, anyone could flood the network with junk for free. Fees prevent spam and price the network's resources.

Validator. A computer on the network that checks transactions follow the rules and then adds them to the ledger, a bit like a notary.

Finality. The point at which a transaction can no longer be reversed.

Stablecoin. A token pegged to a fiat currency. USDC is Circle's dollar stablecoin, fully reserved and redeemable 1:1.

The choices and what they cost

1. USDC is the gas token

On Arc you pay fees in USDC, which means in dollars. You don't first have to buy, hold and account for a volatile native coin.

For a payments company, using Ethereum means taking on risk that has nothing to do with its business. To move one dollar it has to hold ETH. ETH's price moves around, so the dollar cost of fees changes from hour to hour. Holding a speculative asset also pulls the finance team into new accounting rules, tax events and compliance reviews. On Arc a transaction costs a predictable fraction of a dollar, and the treasury never holds anything except dollars.

For ten years the industry argued that a volatile native token was essential, because it pays validators and secures the network. Arc takes the opposite position: for payments, that token is the thing holding adoption back, so the chain has to secure itself some other way.

That raises the obvious question. Without a native gas token, what secures the chain? Usually validators lock up tokens as collateral and lose them if they misbehave. Arc skips that for now and relies on trusted institutions (see choice 5), with a separate ARC token planned for later. Paying gas in dollars makes payments easy, and it forces a completely different security model.

2. Malachite consensus and deterministic finality

Arc's consensus engine is Malachite, a BFT (Byzantine Fault Tolerant) protocol from the Tendermint family. Validators vote on each block in rounds, and once more than two-thirds approve, the block is final.

Bitcoin-style chains never make a transaction strictly final. Reversing it just gets exponentially more expensive as more blocks are added on top. When an exchange asks you to wait for confirmations, it's waiting for a reversal to become implausible, which isn't the same as impossible. Implausible is fine for a trade. A risk committee won't accept it for a $50 million settlement. BFT gives a yes-or-no answer instead: settled, in under a second.

Traditional finance runs on guarantees and not probabilities. DTCC, one of Arc's validators, exists to guarantee settlement. Picking BFT means building for that world and not asking it to adapt.

The cost is decentralization. Every validator has to vote on every block, so the set has to stay small, in the dozens and not the thousands. Ethereum is secure because it has far too many validators to collude. Arc is secure because a handful of institutions have reputations to lose. Arc made that trade knowingly.

3. Consensus separated from execution

Arc splits ordering transactions (consensus, handled by Malachite) from running them (execution, in an EVM). The two are independent layers.

A monolithic chain bundles both, so you upgrade all of it or none of it. With the two apart, you can improve execution without touching consensus, or replace consensus without rewriting every contract. For infrastructure meant to run for decades, that's about maintenance more than elegance, and most chain architects now build this way.

4. EVM compatibility

Arc runs the Ethereum Virtual Machine. Solidity contracts deploy without changes, and Foundry, Hardhat, MetaMask and wagmi work as they are.

Solana's runtime is faster, and Move is arguably a safer language for representing assets. Circle picked the EVM anyway, for its network effects: the most developers, the most audited contract libraries, the deepest tooling and the most people who can review the code. A faster VM with a tenth of the developers is a worse bet.

The EVM has become the default in the way x86 became the default instruction set. It isn't the best design, but its ecosystem makes it hard to leave. New chains now compete on consensus, economics and distribution, and mostly leave the VM alone. Arc fits that pattern.

5. Permissioned validators: PoA now, PoS later

Arc runs Proof of Authority. A fixed, known set of institutions produces blocks, chosen for their identity and reputation and not for staked capital. Circle says the roadmap moves to Proof of Stake later. Validating is permissioned, but deploying isn't. Anyone can ship a contract without asking.

A regulated institution generally can't process payments through infrastructure run by anonymous parties it can't identify or hold to account. That's usually a legal requirement. With PoA the validators are answerable, which is what lets a bank's compliance team sign off.

There's good reason for scepticism, though. Permissioned enterprise blockchains have a poor record. Quorum, Corda, Hyperledger and a series of bank consortium chains launched with fanfare and went nowhere. They were closed networks with no users, no liquidity and no reason for outsiders to build on them.

Arc is different in three ways. It's public, so anyone can read it, transact on it and deploy to it. It launches with a stablecoin that already has real usage, so it doesn't have to create demand from nothing. And it connects to existing ecosystems instead of walling itself off. Whether that's enough remains to be seen.

6. Opt-in privacy on the roadmap

Arc plans selective confidentiality: transaction details hidden from the public but open to authorized parties for audit.

Full transparency gets in the way of commercial payments. If payroll runs on a public chain, every competitor can read every salary. If supplier payments are public, so are the terms you negotiated. Transparency suits a monetary base layer and works against a business running on top of one. Any chain that wants real commercial volume has to solve this.

7. Stablecoin FX and tokenized assets from day one

Arc launched with StableFX, an engine for round-the-clock exchange between fully reserved stablecoins. It supports more than 20 fiat stablecoins, including USDC and EURC, gives direct access to tokenized funds such as BlackRock's BUIDL and Circle's USYC, and has lending through Aave and Morpho. All of it was live from the first block.

Most chains launch empty and spend years trying to attract liquidity. Arc launched with an FX market, money-market instruments and lending already in place, which tells you it was built for known customers with known requirements.

The FX part deserves a closer look. Foreign exchange trades trillions of dollars a day, and it still closes at weekends and settles slowly through correspondent banks. An on-chain FX engine between fully reserved stablecoins that never closes could turn out to be one of the most commercially important parts of this launch.

What it changes for you

Retail traders

Directly, not much. Arc isn't built for you. There's no memecoin culture, no launchpad scene and no free-for-all listing. The validators are a dozen financial institutions, and nothing here is designed for speculative trading.

What reaches you is indirect. USDC liquidity gets deeper on every chain Arc bridges to, and stablecoin FX that never closes makes switching currencies a built-in feature and not a bridge you hope works.

One thing to watch is whether ARC becomes a publicly tradable token. Circle raised $222M in a presale at a $3B valuation, but there are no confirmed details about a public token.

Builders

Your skills carry over completely. Because Arc is EVM compatible, your Solidity and your toolchain work unchanged, and anyone can deploy.

What's new is gas priced in dollars. You can tell a user the exact cost in dollars before they sign, and it's still that cost when the transaction lands. Fee estimators have always tried to approximate that. Subscriptions, payroll, invoicing and anything else that needs a predictable cost per transaction get simpler.

The catch is that blocks come from a permissioned set of institutions. For payments infrastructure that's a reasonable trade. If you need censorship resistance, it rules Arc out.

DeFi protocols

The opportunity is capital that couldn't reach you before. Aave and Morpho are live at launch next to tokenized treasury products like BUIDL and USYC, inside a compliance setup institutions can clear. For a lending protocol that's a new kind of depositor. For FX or payments, StableFX and its 20-plus stablecoins are core infrastructure.

Institutional money brings institutional process with it: audits, legal review and checks on counterparties. It moves more slowly than most DeFi cycles, and that's the price of the larger pool.

Fintech and payments companies

This is who Arc was built for. Unpredictable fees, slow final settlement and a volatile asset on the balance sheet have stopped most enterprise blockchain payment projects. Arc removes all three by design, and it plugs into Circle's existing stack (CCTP, Gateway and institutional on- and off-ramps), so the fiat side is already handled.

Cross-border settlement is the obvious first use. Correspondent banking takes several days, is hard to see into and stacks fees along the way. On Arc you settle in seconds, in dollars, on rails you can explain to a regulator.

Read the fine print, though. Arc hasn't been reviewed or approved by the NYDFS or any other regulator, and the usual blockchain risks still apply. Institutional validators don't make it a regulated product.

AI agent builders

This gets less attention than it should. Circle is pitching Arc for activity driven by AI agents, and a Circle Agent Stack is part of the launch. An agent making economic decisions needs a fixed number to budget against. "0.003 ETH, dollar value unknown until it executes" is a poor input, because the agent has to model an exchange rate just to decide whether an action is worth it. "$0.01" is a number it can act on.

Deterministic finality matters for the same reason. An agent needs to know whether one action succeeded before it takes the next. With probabilistic finality it has to poll and wait. With deterministic finality it gets an answer in under a second, which makes multi-step workflows practical without complicated retry logic.

Put together, that points toward micropayments between agents, metered API calls settled on-chain and automated treasury operations. It's early, but worth watching.

Solana users

Arc isn't competing with Solana. It describes itself as a settlement hub that connects to Ethereum and Solana through CCTP and Gateway. The split is fairly clean. Institutional dollar flows settle on Arc, and open, high-throughput consumer activity (trading, launches, apps with real users) stays where the users already are.

The effect on Solana is indirect but real. More regulated capital on-chain means deeper USDC liquidity everywhere it bridges to, and it's one more big institutional bet that stablecoins are the main use for public chains. That helps every chain where people transact, Solana included.

The bet

All of these choices point the same way. Dollar gas, deterministic finality, known validators and assets from day one were each picked on purpose. Together they bet that the next wave of on-chain volume will be payments and tokenized assets, and that this wave cares more about predictability and accountability than about anyone being able to validate.

If that's right, Arc becomes core infrastructure, and Circle controls both the rails and the dollar. If it's wrong, Arc stays a well-funded consortium chain with an impressive validator list and modest daily use, which this industry has seen before.

Either way, the most institutionally backed chain launch so far is a stablecoin chain and not a general-purpose one. That says a lot about where institutional money thinks this is going.

None of this changes what building on Solana looks like day to day. Solstack is a set of Solana tools that run from your own wallet. You can mint a token, airdrop to holders, multisend to a list and more, with every fee shown before you sign. The docs overview lists every tool.