zksync bridge

My last bridge took 4 minutes on-chain and 11 minutes end to end, because the slow part was not finality. It was the two checks before I pressed confirm.

The first time, I expected the bridge to be the work. Pick asset, pick amount, sign, wait. What actually mattered was duller: whether the receiving wallet already had ETH on Era, whether the token arrived as the version the trade venue used, and whether I had left enough on L1 to undo a bad route without thinking.

Check these first

• Destination gas. If the receiving address has zero ETH on zkSync Era, the bridge can succeed and still leave you parked. I now keep at least 0.003 ETH there before moving anything meaningful.
• Token form. USDC, USDC.e, bridged ETH, native ETH, and pool-specific wrappers are not interchangeable in practice. The balance screen can look right while the route you wanted is still unavailable.
• Minimum received. I set slippage after looking at the actual route, not from habit. A 0.5% setting is not conservative if the pool is thin and the transaction sits.
• L1 residue. I do not bridge the wallet clean. Leave ETH behind for one more approval, one cancellation attempt, or a second route.
• Exit path. Before entering, know whether you would leave through the canonical route, a liquidity bridge, or by swapping first. That choice changes the asset you should enter with.

The useful mental model is that a zksync bridge decision is not only “move funds to Era.” It is “move the right asset, into the right wallet state, with enough remaining gas to use it immediately.” If that sentence is not true before signing, the bridge may still work, but the plan is unfinished.

What I would tell my earlier self: do a tiny first pass. Not symbolic dust, but enough to test the exact next action. For me that is usually $25 to $50. Bridge it, wait for the balance, then attempt the real swap or LP action without completing it. If the route appears, gas estimates make sense, and the token label matches, send the main amount.

The numbers are less interesting than the order. Check the destination first. Check the asset second. Check the route third. Then bridge. Most mistakes happen because those steps get reversed.