The Question Everyone Asks Wrong: How Does Concordium's PayFi Stack Compare?

A reader asked, under the Open Floor initiative, how Concordium's PayFi stack compares to competitors now that Protocol 10 has shipped. It is the right question asked in the wrong frame.

Almost every answer to it on the internet is a side-by-side feature table that confirms whatever bias the author brought to the exercise. Solana wins on speed. Stellar wins on payment heritage. Ethereum wins on liquidity. Concordium wins on compliance. Next slide.

That is not an answer. That is an argument pretending to be one.

The better question is what a PayFi stack actually requires, in order, from settlement through distribution. Once you have that list, the comparison writes itself, because most chains are not building a stack at all. They are building a general-purpose execution environment and hoping payments emerge as a use case. There is a difference.

One thing to name up front, because the honest version of this answer depends on it. Architectural completeness is not market leadership. A chain can have the best-designed stack in the category and still trail competitors on liquidity, issuer gravity, and merchant distribution.

Concordium today is in exactly that position. The architecture is ahead of the market position. The argument in this piece is about the stack. The closing section is about the gap between the stack and the market.

What a PayFi stack requires

Six layers, in order of dependency.

The first is settlement, and this layer has two dimensions that most comparisons collapse into one. Block time is the user-perceived speed, the wait at point-of-sale, the delay between signing and seeing the transaction confirmed. Finality is the guarantee, the moment past which the transaction cannot be reorganised or reversed.

Everyday PayFi lives on block time. High-value settlement lives on finality. A serious payment chain needs to be competitive on both. This is already where the field begins to thin out, because several large chains have explicitly traded away one or the other in favour of other properties.

The second is native money. Stablecoins that live at the protocol layer rather than inside a smart contract that some auditor hopes is bug-free. This matters because every smart-contract stablecoin is, in the end, a promise by an issuer backed by code quality. Protocol-level tokens remove the code from the promise.

The third is identity. Not a KYC checkbox bolted onto an exchange, but compliance-ready, privacy-preserving, reusable identity that lives at the same layer as the money.

Without this, every stablecoin issuer has to rebuild the same verification stack, every merchant has to trust a different provider, and every regulator has to audit a different architecture. Identity at the protocol is what makes the rest composable.

The fourth is frictionless spend. Users holding stablecoins should be able to pay in stablecoins. They should not need to acquire and hold a separate token to cover gas. This sounds mundane. It is the single largest point of consumer friction in crypto payments and it has defeated almost every payment UX to date.

The fifth is machine payments. AI agents buying services from other agents, APIs charging per call, autonomous economic actors moving value without human involvement. This layer is new. It is also where the next decade of payment volume will land. The architecture needs to support it natively and, critically, without surrendering self-custody to a smart contract intermediary.

The sixth is distribution. Wallets that users actually hold, exchanges where the tokens actually trade, payment processors that actually route volume, developers who actually build, issuers who actually mint.

A stack can be architecturally perfect and commercially irrelevant. The sixth layer is what turns architecture into market position, and it is the layer where most of the interesting chains are strong and where Concordium has the most work left to do.

A serious PayFi stack needs all six. Anything less is a payments chain with marketing copy.

The architecture comparison

Five archetypes cover the competitive field on the first five layers. The sixth layer, distribution, is best handled in prose afterwards, because that is where the market position story lives and a grid flattens what is actually a qualitative judgement.

The first table covers the archetypes closest to Concordium's design philosophy. The second covers those furthest from it but strongest in their own categories.

The table above covers the closest architectural neighbours. The table below covers the chains furthest from Concordium's design philosophy but strongest in their own categories.

The payment-first L1s are the closest philosophical cousins. Stellar and XRPL were built for cross-border settlement. Both handle the first layer well, with fast block times and deterministic finality in the 3–5 second range.

Both have also moved recently. XRPL has added credentials and permissioned domains. Stellar's Soroban gives it a path to x402-style machine payments. These are real additions and they matter.

The architectural point still stands, though. Stellar's fee-bump mechanism can sponsor a fee, which looks superficially similar to P10, but fee-bump is a fee primitive only. The richer control logic, delegation, parental spending limits, agent sponsorship with identity checks, all of that would need to be rebuilt in Soroban as smart contracts.

Identity on both chains is a feature layered in rather than a protocol primitive. Both have begun adding compliance and machine-payment capabilities at higher layers, but the trust model remains assembled from components rather than native to the protocol.

The general-purpose giants, Ethereum, Solana, Polygon, Avalanche, Base, share one architectural commitment that disqualifies them from being PayFi stacks. Everything is a smart contract. Stablecoins are contracts. Identity, where it exists at all, is an off-chain database queried by contracts. Paymaster systems that approximate frictionless spend are contracts. x402 facilitators on these chains are contracts that take custody of agent tokens to manage payments.

Each layer adds an attack surface, a custody question, and a gas cost.

Their settlement profiles are also mixed. Ethereum's probabilistic finality reaches economic certainty on a multi-minute horizon, which is why exchanges wait for confirmations before crediting deposits. Solana has sub-second block times but multi-second finality and a history of outages that have in practice invalidated finality during the incidents. Base and other Ethereum L2s have soft sequencer finality but true finality only when batches settle on Ethereum, introducing a multi-minute to multi-hour tail.

The enterprise-compliance chains, Hedera and Algorand, are the closest architectural neighbours. Both take compliance seriously. Both have strong settlement profiles, fast block times and deterministic finality.

Hedera's governance council and Algorand's institutional DeFi positioning overlap with Concordium's target market. Neither has protocol-level identity with zero-knowledge attribute proofs. Algorand's ASA tokens are closer to protocol-native than ERC-20s but still lack the identity binding. Hedera's token service similarly. They get four of the six layers partially. They do not get layers four or five, or the identity-binding piece of layer three.

The privacy chains are a different category, included because the question of PayFi cannot be separated from the question of transactional privacy.

Zcash, Aleo, Aztec, and Midnight each deliver something at layer three that no general-purpose chain delivers: genuine transactional confidentiality. Each also fails the rest of the stack. Zcash inherits Bitcoin's probabilistic finality, which is the wrong settlement model for payments. It has no identity, no native tokens beyond ZEC, no sponsored transactions. Aleo, Aztec, and Midnight are application-layer privacy environments, largely smart-contract based, with their own versions of the custody problem.

Privacy without a stack is a feature, not an infrastructure.

The institutional-permissioned chains are represented most seriously by Canton. This is a different kind of competitor, and for regulated capital markets it is the most credible one.

Canton offers deterministic settlement, privacy calibrated for institutional workflows, and a governance model that capital markets participants understand. It is genuinely stronger than Concordium on institutional entrenchment, regulated-market privacy, and the kind of deployed capital markets credibility that takes years to build.

Where Concordium separates is on being a public, open, permissionless chain. Canton's architecture is not designed to be a public settlement layer. It is designed to be a network of sub-networks where participants are known and admitted. That is a viable model for a specific market, but it is not PayFi in the public-chain sense. For an issuer choosing between the two, the question is which side of that divide they want to live on.

Distribution: the sixth layer

The architecture conversation is the easy part. Distribution is harder and it is where Concordium has the most work to do.

Exchange liquidity is thin compared to Solana or Ethereum. Stablecoin issuer traction is real but early. Developer mindshare is a fraction of what the general-purpose chains command. Merchant integration is emerging rather than entrenched.

The general-purpose giants, Ethereum and Solana in particular, have an order of magnitude more of all of these than Concordium does. Stellar has anchor networks. XRPL has institutional relationships. Canton has capital markets penetration. That weight matters and it is not waved away by a better architecture.

Concordium's distribution story is not empty. It is specific. The wallet partnerships with Ledger, Bitcoin.com, Safle, and Coin98 give it real distribution in age-verified payment verticals where identity is the differentiator. The Ubyx partnership gives Concordium a seat at the regulated stablecoin clearing table alongside Chainalysis, Fireblocks, Solana, Stellar, Ripple, and Paxos.

These are meaningful footholds. They are not followed by any volume yet.

The P10 turn

This is where the shape of the answer changes. Until March 2026 Concordium had layers one, two, and three. Settlement via ConcordiumBFT with 2-second block times and 4-second finality. Native money via PLTs shipped in P9. Identity via the ZK-native ID layer that has been there since genesis.

What was missing was the consumer UX layer and the machine payment layer. P10 ships both in a single structural change.

Sponsored transactions complete layer four. A merchant, wallet provider, or stablecoin issuer constructs the transaction, sponsors the CCD fee, and hands it to the user for signature. The user signs for the stablecoin spend and nothing else. The CCD dependency is invisible to the spender.

That is what frictionless spend looks like when it is not a workaround.

The same primitive completes layer five, and that is the larger story. x402, the emerging standard for machine-to-machine payments, needs a facilitator that can orchestrate payment flows between agents. On other chains the facilitator is a smart contract that takes custody of the agent's tokens.

Concordium's self-custodial design made that impossible without contradicting its own philosophy. Sponsored transactions solve it. The facilitator sponsors the fee without ever touching the agent's tokens. The agent keeps custody. No contract intermediary. This is the only public-chain architecture that delivers x402-style machine payments without smart-contract custody.

One more detail worth noting. The P10 header structure was deliberately built to accept new capabilities without further structural change. Future layers, delegated spending, programmable compliance triggers, new sponsorship models, can ship as feature flags rather than protocol upgrades.

Layers one through five are not only complete; they are extensible.

Where the stack is not yet complete

Two honest acknowledgements, because the argument depends on both.

The first is privacy. Concordium today delivers layer three at Level 1 privacy, identity privacy with anonymity revocation. Level 2 and Level 3, transactional amount privacy and full confidential transfers, are roadmap, not product.

The Shielding Vision series (available exclusively to Inner Circle members) has argued at length that a serious PayFi stack needs Level 2 as a bridge and Level 3 as a destination, because institutional payment volume will not settle on transparent ledgers. That remains true. Until it ships, the privacy layer of the stack is partial, and competitors with Zcash-style confidentiality have a genuine advantage on that specific axis.

The second is distribution, which the section above already covered. The work to close the sixth layer is adoption work, not engineering work. Different muscle, different clock.

The rest of the stack is complete. The honest summary is that Concordium is the only public L1 where five of the six layers are protocol-native, and the sixth is the one solved by adoption rather than engineering.

The answer

The question was never which chain has the best PayFi features. It is which chain was architected as a PayFi stack. After P10, there is one credible public-chain answer on the architecture side.

The market side is still being written.