You hold a token on Ethereum. You want to use it on Arbitrum. What happens next depends on how that token was designed and which infrastructure you use.

Blockchains are isolated systems. Ethereum doesn't know what's happening on Arbitrum. This isolation is what makes them secure, but it creates a coordination problem: if a token has a fixed supply of 1 million, and you want it usable on five chains, something needs to ensure the total never exceeds 1 million.

Several models solve this. Each makes different tradeoffs around trust, speed, cost, and user experience.

Bridging

Bridging moves tokens through a lock-and-mint mechanism. You deposit tokens into a smart contract on the source chain. That contract locks them. A bridge protocol verifies the deposit, then mints an equivalent "wrapped" version on the destination chain. When you return, the process reverses: burn the wrapped token, unlock the original.

Bridge security depends on who validates that deposits happened. LayerZero lets applications configure their own Decentralized Verifier Networks. Wormhole uses 19 validators requiring supermajority consensus. Chainlink CCIP separates verification from execution with a dedicated Risk Management Network.

The output is a wrapped token. Wrapped USDC from one bridge isn't fungible with wrapped USDC from another. This fragments liquidity across incompatible versions.

Native minting

Native minting deploys the token contract directly on each supported chain. Users don't bridge. They mint on whichever chain they want.

Supply coordination happens at the protocol level. When tokens move between chains, they're burned on the source and minted on the destination. Or the protocol enforces per-chain caps. Standards like LayerZero's OFT and Wormhole's NTT make this easier to implement.

User experience is one step: connect wallet, mint. No wrapped tokens, no fragmentation. Same canonical token everywhere.

Liquidity networks

Liquidity networks use market makers who hold inventory on multiple chains. You send tokens to the network on Chain A, they send you native tokens on Chain B. No wrapping, no minting. Just a swap of equivalent value across chains.

The network's liquidity providers take on the cross-chain complexity. You receive native assets, not wrapped representations.

The tradeoff is liquidity dependence. If pools are thin, slippage increases. And you're trusting the network's economic security model rather than cryptographic proofs.

Atomic swaps

Atomic swaps enable direct peer-to-peer exchange using hash time-locked contracts (HTLCs). Both parties lock funds in smart contracts. A cryptographic secret unlocks both sides. Either the swap completes entirely or both parties get refunds.

No intermediary. No wrapped tokens. Fully trustless.

But atomic swaps require compatible chains, willing counterparties, and technical setup. They work for specific swap scenarios, not general token movement. Adoption remains limited.

Intent-based systems

Intent systems let users declare what they want without specifying how to get there. You broadcast: "I want 100 USDC on Arbitrum." Solvers compete to fill your order using whatever method works, whether that's bridges, liquidity networks, or their own inventory.

In practice, many intent systems rely heavily on liquidity networks and inventory held by solvers.

The complexity shifts from users to professional solvers who optimize routes, absorb finality risk, and front capital.

Users get simple UX and often better pricing. The tradeoff is trusting solver infrastructure and settlement guarantees.

When each makes sense

Bridging works for any token but creates wrapped versions and fragments liquidity. Native minting eliminates fragmentation but requires protocol-level support. Liquidity networks handle native assets but depend on pool depth. Atomic swaps are trustless but impractical at scale. Intent systems abstract everything but add solver dependencies.

No single method dominates. The right choice depends on what token you're moving, where you're starting, and what tradeoffs you'll accept.

For example, the Kelp dApp lets users mint rsETH natively on supported L2s or bridge between chains, offering both paths depending on starting position.

Cross-chain infrastructure is evolving fast. These methods will likely blur as chain abstraction matures and intent-based systems route automatically across all of them. But for now, understanding how each works helps you make better decisions about where your assets live and how they get there.

Disclaimer: This is not financial advice. Always DYOR and understand the risks involved before depositing into any DeFi protocol.