How do non-custodial swap services move assets between blockchains

05 September 2026

Non-custodial swap services move assets between blockchains by using atomic swaps, hash-locked contracts, or aggregators that route through multiple liquidity pools, all without ever taking custody of your funds. The core mechanism relies on cryptographic conditions that ensure either both sides of the trade execute or neither does.

The atomic swap mechanism

The most common method is a hash-locked atomic swap. Here is how it works with a simple two-chain example:

1. You send your asset (say, Bitcoin) to a smart contract on the source chain. The contract locks it behind a cryptographic hash. You know the secret that produces that hash; the counterparty does not.

2. The counterparty sends their asset (say, Ethereum) to a similar contract on the destination chain, locked behind the same hash. They set a time limit: if the swap is not completed within that window, the funds return to them.

3. You reveal the secret to claim the Ethereum on the destination chain. That act of revealing makes the secret public, so the counterparty can now use it to claim your Bitcoin on the source chain.

If either party fails to act, the timelock expires and both assets are returned to their original owners. No one loses funds. No one holds them.

Liquidity pools and the role of the exchanger

In practice, many non-custodial swap services do not rely on a single counterparty per trade. Instead, they pool liquidity from many users. The exchanger maintains a set of smart contracts on each supported blockchain. These contracts hold reserves of various assets.

When you initiate a swap, the exchanger's system finds a route through these pools. It might convert your asset to a stablecoin on the source chain, then move value to the destination chain via a cross-chain messaging protocol, and finally convert to your target asset. The exchanger never touches your private keys. The smart contracts handle every step.

The system uses atomic swaps between the pools themselves. The exchange is settled only when all legs of the route succeed. If any leg fails, the entire transaction is reversed.

Cross-chain messaging and wrapped assets

Some services use a lightweight cross-chain messaging layer. The source chain contract locks your asset and emits a message that the destination chain contract can verify. The destination contract then releases the equivalent asset, often a wrapped version of the original. The wrapping is a token that represents the locked asset and can be redeemed later.

This method is faster than a pure atomic swap but introduces a trust assumption: the messaging layer must be secure and honest. Non-custodial services minimise this by using multiple independent relayers or threshold signatures.

Settlement without a bridge

The key difference from a bridge is that no centralised custodian or multi-signature wallet holds your funds during the swap. With a bridge, you deposit assets into a contract that is controlled by a group of signers. With a non-custodial swap, your assets are locked in a contract that only you can unlock, using the cryptographic secret.

The settlement is final when both chains confirm the transactions. Because the contracts are deterministic, the outcome is guaranteed by the blockchain's own consensus, not by the honesty of a third party.

What to read next

If you want the full picture of how these methods compare to other approaches, see the hub page "Swapping crypto across chains". It covers the trade-offs between speed, cost, and trust that each method brings.

Back to dead tokens