What Is LayerZero? The Omnichain Messaging Protocol Behind 75% of Bridge Volume, Explained

Most people meet LayerZero without knowing it — they bridge some USDC, mint an "omnichain" token, or move USDT between chains, and the plumbing underneath is LayerZero. It is an omnichain messaging protocol: infrastructure that lets a smart contract on one blockchain send a verified message to a contract on another, and by mid-2026 it carries roughly 75% of all cross-chain bridge volume across more than 90 chains. That makes it less a "bridge" and more the default interoperability standard — the rail that other bridges, stablecoins, and thousands of apps run on. This guide covers what it actually is, the DVN security model that defines it, the ZRO token and its stalled fee-switch burn, the numbers, the ecosystem, how to use it, and the honest risks — including the $292M exploit that put its whole design philosophy on trial.

What LayerZero is — and who built it

LayerZero is a messaging primitive, not a destination. When people say "bridge on LayerZero," they usually mean an app like Stargate; LayerZero itself is the layer those apps call to pass data between chains. The unit of work is a message — it can carry a token transfer, a governance vote, a swap instruction, or any arbitrary bytes — and the protocol's job is to get that message from a contract on chain A to a contract on chain B with a verification guarantee the receiving app trusts.

The company behind it, LayerZero Labs, was co-founded by Bryan Pellegrino (CEO), Ryan Zarick (CTO), and Caleb Banister. It is one of the best-funded projects in the sector: a $135M Series A in March 2022 at a $1 billion valuation, followed by a $120M Series B in April 2023 that valued it at $3 billion, led by a16z crypto with a long roster of backers including Sequoia Capital, Christie's, Samsung Next, OKX Ventures, and Circle Ventures. That war chest matters — building and maintaining integrations across 90-plus heterogeneous chains is expensive, and it is a big part of why LayerZero's coverage is wider than any competitor's.

The design goal has been consistent since the v1 whitepaper: give applications an interface where sending a cross-chain message is as simple as an ordinary contract call, while pushing the choice of how much to trust that message down to the application itself. That second half is the interesting part.

How LayerZero works: endpoints, DVNs, and the security stack

The current architecture is LayerZero V2, and it has three moving parts.

1. Endpoints. An immutable Endpoint contract is deployed on every supported chain. It is the single, unchanging entry and exit point — your app talks to the Endpoint, and because it is immutable, the interface can't be rug-pulled or upgraded out from under you. Everything else is pluggable around it.

2. DVNs (Decentralized Verifier Networks). This is LayerZero's defining idea. A DVN is an independent party whose job is to watch the source chain and attest that a given message really was emitted — confirming the payloadHash matches. Instead of one fixed validator set (the Wormhole/CCIP approach), LayerZero exposes a marketplace of 30-plus DVNs using different verification methods: zk proofs, middlechains, consortium signers, and third-party bridges. Operators include Google Cloud, Polyhedra, Nethermind, and Chainlink itself.

3. Executors. Once the required DVNs have attested, an Executor submits the transaction on the destination chain and pays that chain's gas, delivering the message to the target contract. This is what lets users pay once on the source chain and have the message "just arrive."

The security model that ties this together is the X-of-Y-of-N configuration. Each application defines its own security stack: X required DVNs that must all sign, plus any Y of a further set of N optional DVNs. The canonical example is "2-of-3-of-5" — two named verifiers must sign, and three of five others must agree. Raise the numbers and you get more redundancy and higher cost; lower them and you get cheaper, faster, riskier messages. Crucially, the app chooses. LayerZero doesn't dictate a single global trust assumption; it hands the dial to developers. Remember that sentence — it explains both LayerZero's flexibility and its worst incident.

In 2024 LayerZero went further, partnering with EigenLayer on a CryptoEconomic DVN Framework that lets a DVN back its attestations with staked assets (ZRO, EIGEN, or ETH) under a stake / verify / veto / slash model — so a DVN that signs a fraudulent message can be economically punished, not just removed. It is an attempt to put money, not just reputation, behind each signature.

The ZRO token and the fee-switch burn

LayerZero launched its token, ZRO, in mid-2024. The supply is fixed at 1 billion and non-inflationary. Allocation is roughly community 38.3%, strategic partners 32.2%, core contributors 25.5%, and repurchased tokens 4%, with a vesting schedule running into 2027; circulating supply is still only a fraction of the total (roughly a quarter to a third) in mid-2026, so unlocks remain a real overhang to watch.

For its first year ZRO was mostly a governance token, and the interesting 2025-26 story is that the community has repeatedly tried to change that — and repeatedly failed. The flashpoint is the long-debated fee switch: the mechanism that would give the token cash flow. It is contentious, and passing requires a steep quorum — roughly 40% of circulating ZRO — that the vote keeps missing even when the overwhelming majority of cast ballots say yes. The LayerZero Foundation has run the referendum four times — December 2024, June 2025, December 2025, and June 2026 — and every one has ended the same way: quorum not met, protocol fee stays off. Referendum #3 (December 2025), for instance, drew over 97% support among votes cast yet still fell short of quorum.

So the "deflationary ZRO" narrative remains a proposal, not a reality as of mid-2026. If a future vote ever clears quorum, the design is well specified: a protocol fee, up to the cost of verifying and executing each message, would be collected across LayerZero messaging and the project's own Zero blockchain, pooled, converted into ZRO at market rate, and sent to a burn address — turning ZRO into a buyback-and-burn asset whose burn rate scales with usage. But until quorum is actually reached, ZRO's on-chain value accrual is governance rights plus the separate Stargate buyback (below), not a live protocol-fee burn.

The numbers: scale, volume, and market position

The headline metric is dominance. As of late 2025, LayerZero powered about 75% of all cross-chain bridge volume, averaging roughly $293 million in daily transfers. It is deployed on more than 90 chains (LayerZero markets even higher figures counting every deployment), spanning not just EVM L2s but Solana, Aptos, Sui, and TON — the widest coverage in the category. Over its life it has relayed 200 million-plus messages.

Positioning matters here. Cross-chain interoperability in 2026 is a genuine market with several credible players — Chainlink CCIP, Wormhole, and Axelar among them — and by TVL the field is more balanced than the volume figure suggests, with the top five protocols controlling roughly 58% of bridge TVL collectively. LayerZero's edge is throughput and integration breadth rather than locked capital: it wins by being the messaging layer the most apps have already chosen to build on. More than 50,000 cross-chain applications were reported building on it by early 2026.

The ecosystem: Stargate, OFTs, and USDT0

Two things sit on top of LayerZero that most users actually touch.

Stargate Finance is the flagship. It is a pooled-liquidity bridge — deposit into unified pools and move native USDC, USDT, ETH, or BTC across chains — and it is LayerZero's most-used front end. The relationship got closer in 2025: the LayerZero Foundation acquired Stargate in a deal worth about $110 million, which the Stargate DAO voted to accept on August 24, 2025, dissolving its separate governance. Under the deal terms, half of Stargate's top-line revenue funded ZRO buybacks while the other half went to STG/Stargate stakers during a six-month window; after that window closed in February 2026, Stargate's revenue now flows into ZRO buybacks — wiring the ecosystem's busiest app directly into the token. (Note this buyback is separate from, and not conditional on, the still-inactive protocol fee switch.)

OFT — the Omnichain Fungible Token standard — is arguably more important long-term. Instead of the old wrapped-asset model (lock a token on chain A, mint an IOU on chain B, and hope the lockbox holds), an OFT is a single token that burns on the source chain and mints on the destination, moving supply rather than issuing derivatives. More than 730 OFTs now exist, and the marquee examples are heavyweight: USDT0, an omnichain version of Tether's USDT, and PayPal's PYUSD. Collectively OFTs have processed well over $160 billion in cross-chain transfers. That is the quiet reason LayerZero is hard to dislodge — major stablecoin issuers have standardized their multichain presence on it.

How to actually use LayerZero

Most people use LayerZero indirectly. Two paths:

1. As a user, via an app. Go to a front end built on LayerZero — stargate.finance is the obvious one for asset transfers. Connect your wallet, pick source and destination chains and the asset, and confirm. You pay a fee on the source chain that bundles the messaging cost plus destination gas; the DVNs verify and an Executor delivers the tokens on the other side. For OFT tokens (like USDT0), the token's own app or many wallets expose the same cross-chain send natively.
2. As a developer, via the protocol. Build an OApp (omnichain application) or issue an OFT. You inherit the immutable Endpoint on each chain and then — this is the load-bearing step — configure your security stack. Do not accept a weak default. Set a multi-DVN X-of-Y-of-N policy with independent verifiers, because that single choice is the difference between a robust deployment and the next cautionary tale.

Practical note for everyone: cross-chain sends are near-instant but not always, and messages can occasionally stall waiting on DVN or Executor delivery. Use official front ends, verify contract addresses, and treat any "your bridge is stuck, click here" support DM as a scam.

Risks and what to watch: the KelpDAO exploit

LayerZero's honest risk profile is defined by one word: configuration. The protocol's core contracts have never been exploited — but that is not the same as saying nothing has gone wrong.

On April 18, 2026, attackers linked to North Korea's Lazarus Group (specifically the TraderTraitor cluster) stole roughly $292 million — about 116,500 rsETH — from KelpDAO's LayerZero bridge. Read the mechanism carefully, because it is the whole lesson. This was not a smart-contract hack. The attackers went after off-chain DVN infrastructure: they compromised RPC nodes the DVN relied on, swapped node binaries, DDoS'd the healthy RPCs to force a failover onto the poisoned ones, and used that foothold to forge a message the DVN then attested. The forged message minted rsETH out of thin air on the destination.

Why did nothing catch it? Because KelpDAO was running a 1-of-1 DVN configuration — a single verifier, with LayerZero Labs as the sole signer. There was no independent second attestor to reject the forgery. This is precisely the risky end of the dial LayerZero hands to developers, and it directly contradicts the multi-DVN redundancy the protocol has always recommended. A 2-of-2 or 3-of-3 stack would very likely have stopped it.

The aftermath was a public blame fight — KelpDAO argued LayerZero had approved and effectively defaulted them into the vulnerable setup; LayerZero argued the customer chose a single-verifier config against guidance. Both things can be partly true, and that tension is the real risk to understand: LayerZero's flexibility shifts meaningful security responsibility onto app developers, and not all of them use it wisely. Concretely, LayerZero says its DVN will no longer sign or attest for any app on a 1/1 configuration, is reaching out to remaining single-DVN projects to migrate them onto redundant multi-DVN setups, and runs one of the sector's largest bug bounties (up to $15M on Immunefi). KelpDAO, for its part, moved rsETH to Chainlink CCIP.

Other things to watch: token unlocks running into 2027, competitive pressure from CCIP and Wormhole (both of which lean on a single, arguably simpler-to-reason-about trust model), and the plain fact that concentrating ~75% of bridge volume in one protocol makes it a systemic target.

Bottom line

LayerZero won the interoperability race the way TCP/IP won networking — by becoming the layer everyone else builds on, then getting the biggest issuers (Tether, PayPal) to standardize on its token format. Its defining choice, configurable per-app security via DVNs, is genuinely powerful: it lets a serious team build something more robust than any fixed-validator bridge. But that same choice is a loaded gun for teams that pick a weak setup, as the $292M KelpDAO exploit showed in the most expensive way possible. With Stargate folded back in and its revenue funding ZRO buybacks — though the headline fee switch has yet to clear a governance quorum — the token's economics are still a work in progress, while the protocol's contracts remain unbroken. The takeaway for users is simple: LayerZero is the dominant, best-integrated cross-chain rail in 2026, and its safety depends almost entirely on how the specific app you're using has configured it. Ask which DVNs verify your transfer. If the answer is "just one," walk away.

Related guides

• Stargate review
• USDT0 review
• Wormhole review
• Kelp (KelpDAO) review
• Cross-chain crypto bridges

Not financial advice. ZRO is volatile and cross-chain bridging carries real smart-contract and configuration risk — verify every contract address and DVN setup on official channels before moving funds.