ZKsync Era represents one of the most technically ambitious projects in the ZK rollup space. Developed by Matter Labs (founded in 2018, headquartered in Berlin), ZKsync was one of the first production ZK rollups on Ethereum mainnet — and the first to launch a ZK-EVM (a ZK prover that can verify arbitrary Ethereum smart contract execution, not just token transfers). The June 2024 ZKSYNC token airdrop generated significant controversy: the distribution was perceived as favoring Sybil addresses and leaving out many legitimate early users, becoming one of the most criticized airdrops in recent memory. Despite the controversy, ZKsync Era remains one of the most technically sophisticated L2s, and its ZK Stack (the modular framework for building custom ZK hyperchains) positions it as both an L2 and an L2-creation platform.
Technical Architecture
The protocol is built around the following components.
ZK-EVM: The Core Innovation
A ZK-EVM is a system that can:
- Execute any Ethereum smart contract (EVM-compatible)
- Generate a zero-knowledge proof that the execution was correct
- Verify that proof on Ethereum L1 in a single transaction
Why this is hard:
- The EVM was designed for straightforward execution; its design is not “ZK-friendly” (many operations are expensive to prove in ZK circuits)
- Early ZK rollup systems (StarkEx, ZKsync V1, Loopring) only handled specific operations (token transfers, specific DEX logic)
- Building a general ZK-EVM means proving any EVM opcode in ZK — each opcode requires a specific ZK circuit
ZK-EVM approaches (Vitalik’s taxonomy):
- Type 1 (fully Ethereum-equivalent): Proves actual Ethereum blocks; most compatible; proof generation too slow for production
- Type 2 (EVM-equivalent): Proves EVM execution semantics; full EIP compatibility; slightly modified data structures
- Type 3 (almost EVM-equivalent): Most EVM compatible; few compromises for prover efficiency
- Type 4 (high-level language equivalent): Compile Solidity to ZK-friendly LLVM IR; fast proofs; not opcode-compatible
ZKsync Era’s approach: Roughly Type 4 — Solidity/Vyper compilered to its own intermediate “LLVM IR for ZK,” then proven. This means ZKsync is language-compatible but not bytecode-compatible with Ethereum; contracts must be recompiled for ZKsync (not just deployed).
Boojum: The ZK Proving System
ZKsync’s current proving system is Boojum (launched 2023):
- Based on PLONK + FRI (no trusted setup needed; transparent)
- Significant GPU acceleration for proof generation
- Recursive proving: Aggregate multiple L2 blocks into one L1 proof
- Performance: Can generate proofs faster than predecessor “SNARK-based” system
Prior system: ZKsync originally used a Groth16-based proving system (required trusted setup). Boojum migration removed the trusted setup requirement.
Native Account Abstraction
ZKsync Era’s most user-friendly feature: every account is a smart account by default.
On Ethereum:
- EOAs (Externally Owned Accounts) = private key controlled; no programmability
- Smart contract accounts = programmable but require ERC-4337 bundler infrastructure
On ZKsync Era:
- All accounts (including “simple” ones) are smart contract accounts by default
- No distinction between EOA and contract account at the protocol level
- Benefits: Paymasters (gas sponsorship), multi-sig natively, session keys, social recovery — all without additional infrastructure
This is a fundamentally different architecture than Ethereum’s ERC-4337 overlay and represents a cleaner implementation of account abstraction.
ZK Stack: Hyperchains
ZK Stack is Matter Labs’ framework for building custom ZK-powered chains:
What ZK Stack Provides
- A deployable ZK rollup stack based on ZKsync Era code
- Ability to create a Hyperchain — a custom ZK rollup that shares ZKsync Era’s security model
- Configurable: Custom token for gas, custom governance, custom data availability (Ethereum calldata, Eigen DA, custom DA)
- Interoperability: Native cross-Hyperchain bridges through shared ZK proof aggregation
Notable ZK Stack Deployments
| Project | Use Case |
|---|---|
| Cronos zkEVM | Crypto.com’s ZK chain |
| Sophon | Gaming and entertainment chain |
| Treasure Chain | Gaming ecosystem |
| Playnance | Gaming-focused ZK chain |
| Multiple others | Enterprise, DeFi, gaming |
ZK Stack vs. OP Stack
OP Stack (Optimism): Optimistic rollup framework; Base, Mode, Metal, and dozens of others
ZK Stack (Matter Labs): ZK rollup framework; fewer but growing deployments
Key difference: OP Stack chains inherit Optimism’s optimistic security model (7-day fraud window); ZK Stack chains inherit ZKsync’s ZK security model (immediate finality once ZK proof submitted).
The ZKSYNC Airdrop Controversy
Here is what happened.
The Airdrop
On June 17, 2024, Matter Labs announced the ZKSYNC token and airdrop:
- Total supply: 21 billion ZKSYNC
- Airdrop allocation: 17.5% (3.675 billion ZKSYNC) to early adopters
- Eligible wallets: ~695,000 wallets
Why the Community Was Angry
The Sybil problem:
- ZKsync’s airdrop criteria heavily weighted transaction count and fee paid
- Professional Sybil farmers (who created hundreds of wallets and farmed minimum activity) received the same or more ZKSYNC than genuine long-term users
- Many legitimate users who used ZKsync once early (2021–2022, when it was genuinely experimental) received nothing or tiny allocations
The “too many wallets” problem:
- 695,000 eligible wallets is one of the largest airdrop sets ever
- By contrast, Arbitrum distributed to ~350K wallets; Optimism phased distributions more carefully
- Spreading 17.5% of supply across 695K wallets = very small individual allocations for most
Comparison concern:
- Starknet’s STRK airdrop had similar backlash but ZKsync received more critique for perceived arbitrary criteria
- Heavy early Layer Zero (ZKsync V1) users who pre-dated ZKsync Era found themselves ineligible
Market reaction: ZKSYNC launched at ~$0.35, dropped to ~$0.12–0.15 within months; total market cap quickly dropped below initial launch price → reflecting the airdrop-dump pattern common to large-allocation airdrops.
ZKsync Competitive Position
The following sections cover this in detail.
vs. Starknet
| Factor | ZKsync Era | Starknet |
|---|---|---|
| VM | ZK-EVM (Type 4) | Cairo VM (non-EVM) |
| Language | Solidity/Vyper → recompile | Cairo (custom language) |
| Proving system | Boojum (PLONK+FRI) | STARK |
| Account abstraction | Native | Native |
| Token | ZKSYNC | STRK |
| TVL (2025) | ~$200–400M | ~$300–600M |
ZKsync vs. Starknet: ZKsync is more EVM-compatible (Solidity developers can port code); Starknet offers better ZK-native performance but requires learning Cairo.
vs. Polygon zkEVM
| Factor | ZKsync Era | Polygon zkEVM |
|---|---|---|
| EVM compatibility | Type 4 (recompile required) | Type 2/3 (bytecode-level compatible) |
| Ecosystem | Smaller but native-first | Backed by Polygon ecosystem |
| Token | ZKSYNC | POL/MATIC |
Polygon zkEVM is more bytecode-compatible (no recompile needed); ZKsync Era offers better developer UX for native ZKsync-first development.
How to Use ZKsync Era
Bridge: Native ZKsync bridge at portal.zksync.io; also via layerswap, Orbiter Finance
Wallet: MetaMask (add ZKsync network); native ZKsync wallet; Argent (native account abstraction)
DeFi: SyncSwap (main DEX), Mute.io, Maverick Protocol (concentrated liquidity), zkSync DeFi ecosystem
ZKSYNC token: Available on Binance, OKX, and other exchanges.
Hardware security: ZKSYNC is EVM-compatible; store with Ledger via MetaMask.
Related Terms
Sources
Groth, J. (2016). On the Size of Pairing-Based Non-Interactive Arguments. EUROCRYPT 2016.
Ben-Sasson, E., Bentov, I., Horesh, Y., & Riabzev, M. (2019). Scalable Zero Knowledge with No Trusted Setup. CRYPTO 2019.
Buterin, V. (2022). The Different Types of ZK-EVMs. Vitalik Buterin’s Blog.
Gabizon, A., Williamson, Z.J., & Ciobotaru, O. (2019). PLONK: Permutations Over Lagrange-Bases for Oecumenical Noninteractive Arguments of Knowledge. IACR ePrint 2019/953.
Wahrstätter, A., Lande, J., Schwarzwälder, D., & Knottenbelt, W. (2023). Blockchain Scalability and the Fragmentation of the L2 Ecosystem. ACM Distributed Ledger Technologies: Research and Practice.