Tokens

Token identity, L1↔L2 mapping, bridge asset IDs, and chain token facts for ETH, base token, and ERC-20s.

At a Glance

Resource: sdk.tokens
Capabilities: resolve tokens, map L1⇄L2 addresses, compute assetId, detect base token, WETH helpers, predict bridged addresses.
Auto-handling: ETH aliases (ETH_ADDRESS, FORMAL_ETH_ADDRESS) and L2 base-token alias are normalized for you.
Error style: Throwing methods; wrap in try/catch or use higher-level try* callers.

Import

import { createPublicClient, createWalletClient, http } from 'viem';
import { privateKeyToAccount } from 'viem/accounts';
import { createViemClient, createViemSdk } from '@matterlabs/zksync-js/viem';

const account = privateKeyToAccount(process.env.PRIVATE_KEY as `0x${string}`);

const l1 = createPublicClient({ transport: http(process.env.L1_RPC!) });
const l2 = createPublicClient({ transport: http(process.env.L2_RPC!) });
const l1Wallet = createWalletClient({ chain: l1Chain, account, transport: http(process.env.L1_RPC!) });
const l2Wallet = createWalletClient({ chain: l2Chain, account, transport: http(process.env.L2_RPC!) });

const client = createViemClient({ l1, l2, l1Wallet, l2Wallet });
const sdk = createViemSdk(client);
// sdk.tokens → TokensResource

Quick Start

Resolve a token by L1 address and fetch its L2 counterpart + bridge metadata:

const tokenAddress = '0xYourToken';
const token = await sdk.tokens.resolve(tokenAddress);
/*
{
  kind: 'eth' | 'base' | 'erc20',
  l1: Address,
  l2: Address,
  assetId: Hex,
  originChainId: bigint,
  isChainEthBased: boolean,
  baseTokenAssetId: Hex,
  wethL1: Address,
  wethL2: Address,
}
*/

Map addresses directly:

const tokenAddress = '0xYourToken';
const l2Addr = await sdk.tokens.toL2Address(tokenAddress);
const l1Addr = await sdk.tokens.toL1Address(l2Addr);

Compute bridge identifiers:

const tokenAddress = '0xYourToken';
const assetId = await sdk.tokens.assetIdOfL1(tokenAddress);
const backL2 = await sdk.tokens.l2TokenFromAssetId(assetId);

Method Reference

`resolve(ref: Address | TokenRef, opts?: { chain?: 'l1' | 'l2' }) → Promise<ResolvedToken>`

Resolve a token reference into full metadata (kind, addresses, assetId, chain facts).

L1↔L2 Mapping

toL2Address(l1Token: Address) → Promise<Address> — returns L2 token; base token → L2_BASE_TOKEN_ADDRESS, ETH aliases normalized.
toL1Address(l2Token: Address) → Promise<Address> — returns L1 token; ETH alias normalized.

Bridge Identity

assetIdOfL1(l1Token: Address) → Promise<Hex>
assetIdOfL2(l2Token: Address) → Promise<Hex>
l2TokenFromAssetId(assetId: Hex) → Promise<Address>
l1TokenFromAssetId(assetId: Hex) → Promise<Address>
originChainId(assetId: Hex) → Promise<bigint>

Chain Token Facts

baseTokenAssetId() → Promise<Hex> — cached.
isChainEthBased() → Promise<boolean> — compares base token assetId vs ETH assetId.
wethL1() → Promise<Address> — cached WETH on L1.
wethL2() → Promise<Address> — cached WETH on L2.

Address Prediction

computeL2BridgedAddress({ originChainId, l1Token }) → Promise<Address> — deterministic CREATE2 address for a bridged token; handles ETH alias normalization.

Notes & Pitfalls

Caching: baseTokenAssetId, wethL1, wethL2, and the origin chain id are memoized; repeated calls avoid extra RPC hits.
ETH aliases: Both 0xEeeee… (ETH sentinel) and FORMAL_ETH_ADDRESS are normalized to canonical ETH.
Base token alias: L2_BASE_TOKEN_ADDRESS maps back to the L1 base token via toL1Address.
Error handling: Methods throw typed errors via the adapters’ error handlers. Wrap with try/catch or rely on higher-level try* patterns.

ZKsync SDK