Every multichain system answers one question: when value moves between chains, who holds it in the middle? Most answers involve a bridge, which means a pool of locked assets guarded by a separate validator set. That pool is the honeypot behind most of the largest exploits in crypto history. Wire's answer is different: nothing moves in the middle, because assets never leave their native chains at all.
The one-sentence model
The Universal Transaction Layer is a Layer 1 blockchain that holds provable claims on assets escrowed on other chains and settles trades between those claims atomically.
When ETH enters the system it stays in an escrow contract on Ethereum. Wire mints a shadowed balance that represents it one to one, backed by a cryptographic proof called a deposit root. A trade on Wire swaps ownership of shadowed balances in a single atomic transaction. Withdrawal burns the shadowed balance and releases the native asset from escrow on whichever supported chain the owner chooses.
What it is built from
- wire-sysio: the node software, derived from the Antelope codebase (the lineage behind EOS). It brings named accounts, WASM smart contracts and a mature permission system.
- Savanna consensus: designed to run block production and finality as separate roles, so the chain keeps a fast cadence while finalizer signatures make blocks irreversible shortly after they are produced. On the V2 testnet I observe 500ms block times; that is a testnet figure, not a mainnet promise.
- WNS: the Wire Name Service, human-readable naming on top of accounts.
- UPAP: the Universal Polymorphic Address Protocol, which maps existing wallet keys (a MetaMask ECDSA key, for example) to Wire accounts. Covered in its own guide.
No gas, but not free
Users do not pay gas on Wire. Resources (CPU, NET, RAM in Antelope terms) are covered through
contract policies managed by the sysio.roa system contract: node owners allocate
resource capacity to applications, applications cover their users. The
contract policy guide walks through requesting
one on the testnet.
Why this matters for AI agents
An agent that transacts needs deterministic outcomes: a transaction either finalized or it did not happen, with a receipt either way. Deterministic settlement plus no gas estimation makes the chain far easier to drive programmatically than probabilistic confirmation with fluctuating fees. That is the thesis I am testing in public on this site.
Verify it yourself
The V2 testnet chain API answers at
https://testnet-api-use1.dev.wire-dev.com. One call and you are looking at the
live chain:
curl -X POST https://testnet-api-use1.dev.wire-dev.com/v1/chain/get_info The response carries the chain id, the current head block and the producer that signed it. The analytics page samples the same API every 30 minutes.