---
title: "ZKsync (ZK & ZK Nation)"
url: "https://caper.network/wiki/daos/networks/zksync-era"
updated: 2026-09-07
license: CC-BY-4.0
license_url: "https://creativecommons.org/licenses/by/4.0/"
---

# ZKsync (ZK & ZK Nation)

| Type | Ethereum L2 rollup / network DAO |
| --- | --- |
| Governance token | ZK (governance only — gas is paid in ETH) |
| Governance framework | ZK Nation — three onchain bodies, three governors |
| Onchain governance live | October 2024 |
| Max supply | 21,000,000,000 ZK — enforced by the token contract only since [ZIP-14](https://forum.zknation.io/t/zip-14-upgrade-zk-token-with-permissionless-burn-function/771) executed, December 2025 |
| Supply minted | 14,901,043,049.6 ZK — just under 71% of the ceiling (read 6 September 2026, ZKsync Era block 71,859,407) |
| Builder | Matter Labs (ZK Stack / Elastic Network) |
| Website | [zknation.io](https://zknation.io) |

## Overview

[ZKsync](https://zksync.io) is an Ethereum layer-2 built by Matter Labs on validity (ZK) proofs. ZKsync Era went live on mainnet in March 2023, and the stack has since generalised into the [Elastic Network](https://docs.zksync.io/zksync-network) — a set of ZK-powered chains built with the open-source [ZK Stack](https://docs.zksync.io/zk-stack) that settle to the same shared bridge. Its governance sits with **ZK Nation**, the community of [ZK token](https://zksync.io/token) holders and delegates that controls the protocol contracts.

ZK is close to a pure governance asset: transaction fees on ZKsync are paid in ETH, so ZK's standing protocol role is voting weight over upgrades, the token itself, and governance operations. The one exception so far was the ZKnomics staking pilot, which paid newly minted ZK to holders who staked and delegated, and which has been paused since May 2026 (below). That separation — a governance token that is _not_ the gas token — puts ZKsync alongside other L2 network DAOs the wiki tracks, such as [Arbitrum](/wiki/daos/networks/arbitrum-dao), the [Optimism Collective](/wiki/daos/networks/optimism-collective), [Starknet](/wiki/daos/networks/starknet), and [GnosisDAO](/wiki/daos/networks/gnosisdao).

## The ZK token & distribution

ZK's tokenomics set a ceiling of **21 billion** tokens — not a supply that exists, and until December 2025 not one the contract enforced (see the next section) — with roughly two-thirds (66.7%) earmarked for the community and the remainder split between investors and the Matter Labs team under a one-year lock followed by a linear unlock from June 2025 to June 2028. The token generation event and initial airdrop landed in June 2024.

The community airdrop distributed **3.675 billion ZK (17.5% of supply)** across **695,232 wallets**, based on a snapshot of ZKsync Era and ZKsync Lite activity taken on 24 March 2024 — about 89% weighted to users who transacted and met an activity threshold, and 11% to ecosystem contributors. Details are set out in Matter Labs' [airdrop announcement](https://web.archive.org/web/20251215233918/https://www.zknation.io/blog/zk-token) (archived; the post has since been removed from zknation.io) and [reported](https://www.coindesk.com/tech/2024/06/11/zksyncs-zk-airdrop-is-coming-next-week-heres-what-to-expect) at the time.

## Supply: a ceiling the contract enforces, and 6.1 billion it has never minted

The 21 billion figure quoted in nearly every write-up of ZK is a _ceiling_, not a supply — and until December 2025 it was not a ceiling the token contract knew anything about. Two keyless reads, both re-runnable, separate the numbers.

- **Minted supply.** `totalSupply()` on the ZK token — `0x5A7d6b2F92C77FAD6CCaBd7EE0624E64907Eaf3E`, native to ZKsync Era — returned **14,901,043,049.600285 ZK** at Era block **71,859,407**, read 6 September 2026. That is just under 71% of the ceiling, leaving **6,098,956,950 ZK** that has never been minted.
- **The ceiling.** `maxSupply()` on the same contract returns exactly **21,000,000,000 ZK**.

The gap is not tokens parked somewhere else. ZK's bridged counterpart on Ethereum, `0x66a5cfb2e9c529f14fe6364ad1075df3a649c0a5`, held **17,996,312.2 ZK** at Ethereum block 25,914,925 the same day — roughly 0.12% of the Era figure, nowhere near six billion. The unminted remainder is schedule rather than custody: it comes into existence as the unlock curve and governance-approved programmes call for it, each through a minter the governance system has to authorise. A reader who takes 21 billion as the denominator for a holdings or quorum calculation is using a number about a third larger than the one that exists.

**The ceiling arrived by vote, and it arrived late.** [ZIP-14](https://forum.zknation.io/t/zip-14-upgrade-zk-token-with-permissionless-burn-function/771), authored and sponsored by ScopeLift, upgraded the token to ZKTokenV3 and added three things: a public `burn` any holder may call against their own balance, a role-gated `burnFrom`, and the `maxSupply()` function read above — with `BURNER_ROLE_ADMIN` assigned to the ZKsync Protocol Governor timelock. L2BEAT's governance team [published its FOR vote and its review](https://forum.zknation.io/t/zip-14-upgrade-zk-token-with-permissionless-burn-function/771/5), calling the change minimal, safe and forward-compatible. It [executed on Ethereum](https://etherscan.io/tx/0x2c324b8b11fb6cd69eacbc35f7e0e9a22830bffba63d2a352b82628f0935a3bb) at block 23,967,898, **8 December 2025 11:50:47 UTC**. Before that transaction the ZK contract carried no maximum at all, and “fixed supply of 21 billion” was a claim in a tokenomics document rather than a property of the asset. That both facts are now true on-chain is checkable in one call each, which is the point of stating them this way.

Note also that ZIP-14 travelled as a **ZIP** through the Protocol Governor, not as a Token Program Proposal, even though its subject was minting and burning rights over ZK — the proposal classes below split on which contracts an upgrade touches, and the boundary is not always where the topic suggests.

## The staking pilot, and what it actually measured

Between **9 February and 11 May 2026** ZK Nation ran Season 1 of the [ZKnomics staking pilot](https://forum.zknation.io/t/tpp-12-zknomics-token-staking/766) (TPP-12): holders staked ZK, delegated the voting power that came with it to an active delegate, and were paid newly minted ZK in return. Tally supplied the infrastructure. The stated purpose was to rehearse staking ahead of a decentralised sequencer and to raise _active_ delegated voting power. The programme's own [Season 1 review](https://forum.zknation.io/t/season-1-review-zknomics-staking-pilot-program/992) reports the results first-party: a peak of **355m ZK staked** against a 400m target (about $7m), **4,397 addresses**, **5.3m of an available 10m ZK** actually distributed as rewards, and zero incidents. Active delegated voting power went from 958m ZK at the announcement to 1.163bn at the close — net new delegation of **205m ZK**. Unminted reward tokens were left unminted and the season's capped minter had its minter roles revoked.

**The reply underneath it is the part worth reading.** [Curia checked the programme's headline metric against a second one](https://forum.zknation.io/t/season-1-review-zknomics-staking-pilot-program/992/2), and the two disagree. Roughly **76% of the pilot's delegation reached six addresses**; the top three alone received 209.7m ZK, more than the entire net gain of 205m, so removing them leaves the headline roughly flat. And measured over the same five-proposal rolling window the programme's own “active delegate” filter uses, average proposal participation moved from 957.76m to 968.77m — about one per cent. A programme can move delegated power a great deal and move votes actually cast barely at all. That gap is exactly what [voting power activation](/wiki/dao-governance/concepts/voting/voting-power-activation) and [delegate accountability](/wiki/dao-governance/concepts/voting/delegate-accountability) are about, and it is a useful caution for any DAO planning to buy participation with emissions: the thing rewarded was delegation, and delegation is what grew.

**Season 2 is paused.** The programme admin [announced the pause](https://forum.zknation.io/t/season-1-review-zknomics-staking-pilot-program/992) in the same post, citing alignment with a decentralised sequencer that has slipped to late H2 2026, custodian limitations that kept institutional holders out, and an APR model it wants to reassess before committing to a permanent staking contract. The pilot's capped minters expire on **31 December 2026**.

## Three-body governance

ZK Nation's defining feature is a **three-body structure** — a deliberate departure from the single-house token vote used by most protocol DAOs. Each body checks the others ([governance 101](https://docs.zknation.io/voting-and-delegation/zksync-governance-101)):

- **Token Assembly** — ZK holders who delegate voting power to ZKsync addresses. The Assembly submits and votes on proposals to upgrade the protocol, the token, and the governance system itself.
- **Security Council** — 12 technical experts (auditors, security engineers) who are signers of a security multisig. They [review and actively approve](https://docs.zknation.io/zksync-governance-procedures/schedule-3-zksync-security-council) protocol upgrades passed by the Token Assembly and, facing an imminent threat, can freeze the protocol and push time-sensitive fixes.
- **Guardians** — protectors of the ZK Credo, a minimum of 5 of 8 allocated seats, holding **veto power** to block decisions that clash with the ecosystem's stated values.

Splitting proposing power (Assembly), technical safety (Council), and a values veto (Guardians) is a different answer to the [capture and rushed-upgrade risks](/wiki/daos/networks/arbitrum-dao) that a plain token vote carries — closer in spirit to a constitutional separation of powers than to one-token-one-vote.

## Onchain governors & proposal types

The three bodies act through three onchain [governor contracts](https://docs.zknation.io/zksync-governance-procedures/zksync-governance-procedures-overview) on ZKsync Era, each paired with a timelock and each owning one proposal class:

- **Protocol Governor** executes **ZKsync Improvement Proposals (ZIPs)** — upgrades to the protocol contracts.
- **Token Governor** executes **Token Program Proposals (TPPs)** — proposals that assign minting and burning rights over ZK.
- **GovOps Governor** facilitates **Governance Advisory Proposals (GAPs)** — governance and operations decisions.

Standard upgrades follow a set path: Token Assembly approval → Security Council risk review → timelocked execution. Emergency changes route instead through the **Emergency Upgrade Board**, which accepts approvals from all three governance bodies before an expedited upgrade can execute.

## BORGs: smart contracts wrapping legal entities

The Security Council and Guardians are structured as **BORGs** (blockchain organizations) — legal entities whose behaviour is constrained by onchain rules embedded in their charters, so a multisig cannot simply act outside its mandate. It is the same "cybernetic org" pattern that [Safe](/wiki/daos/infrastructure/safe-dao) multisigs and adjacent tooling have popularised, applied here to give a values veto and a security freeze real legal and onchain teeth rather than social convention alone.

## How Caper relates

ZKsync's Guardians are a _backstop against capture_: a minimum of five of eight allocated seats can veto a decision that clashes with the ecosystem's stated values. Until 29 August 2026 [Caper](/wiki/foundations/what-is-a-caper) had no equivalent, and this section said so — that its answer to a bad decision was the [exit right](/blog/the-exit-right) rather than a veto seat. That contrast no longer holds, because a caper has since acquired a veto of its own.

A [caper proposal](/wiki/governance/proposals) now clears two gates. The ballot is legislative: it decides which typed option carries a mandate. Triggering that option then locks the caper's trailing TWAP as a baseline and opens the market window, and the action executes if and only if the TWAP measured across that window is at or above the baseline. That is a veto in the standard [optimistic-governance](/wiki/dao-governance/concepts/voting/optimistic-governance) sense — but the seat is not appointed, because there is no seat. Anyone holding the token can exercise it, and exercising it means selling hard enough to hold the price under the baseline for a real share of the window. Holding is consent.

Set beside ZK Nation's three bodies, the difference is who pays for saying no. A Guardian veto is free to the Guardian, and its legitimacy rests on the charter it is judged against. A caper's veto costs the vetoer whatever they sell into, and its legitimacy rests on nothing but the price that results. Neither is obviously the better answer: a values veto can block a decision no market would have punished, and a price veto can be outbid by anyone willing to buy the token back up. The [exit right](/blog/the-exit-right) is still there underneath both — but in a caper it is now the last line of defence rather than the first.