A zkEVM (zero-knowledge EVM) is a Layer-2 scaling solution that executes Ethereum smart contracts the same way Ethereum does — using the EVM — but proves their correct execution using zero-knowledge proofs (ZK-SNARKs or ZK-STARKs) rather than optimistic fraud proofs. This combines EVM compatibility (so existing Solidity contracts and tools work) with ZK rollup security (instant finality, no withdrawal delay). Building a zkEVM is widely considered one of the most technically challenging problems in blockchain engineering.
How It Works
Background: The Problem
Optimistic rollups (Arbitrum, Optimism) execute EVM transactions off-chain but require a 7-day challenge period before withdrawals. ZK rollups provide instant cryptographic proof of correctness — but until 2022, ZK proofs could not handle the complex, general computation of the EVM (only specific applications like token swaps or simple transfers).
How zkEVMs Work
- Transactions are submitted to the zkEVM sequencer
- The sequencer executes transactions using EVM-compatible logic
- A prover generates a ZK proof (ZK-SNARK or STARK) attesting: “I correctly executed these transactions starting from state A and ending at state B”
- The proof is posted to Ethereum’s L1, where a small smart contract verifies it
- State updates are final as soon as the proof is verified on Ethereum (minutes to hours, depending on implementation)
The Key Challenge
EVM opcodes are complex and variable. Proving arbitrary EVM execution requires either:
- Modifying the EVM spec to be more ZK-friendly (Type 2 or Type 3 zkEVM — slight EVM incompatibility)
- Proving native EVM execution exactly (Type 1 zkEVM — maximum compatibility, much harder to prove)
Vitalik’s zkEVM Classification
- Type 1: Fully Ethereum-equivalent (Scroll works toward this)
- Type 2: EVM-equivalent but different hash functions (Polygon zkEVM)
- Type 3: Almost EVM-equivalent (minor differences)
- Type 4: High-level language equivalent (compile Solidity to a ZK-friendly bytecode) — fastest proofs, least compatible (zkSync Era, StarkNet)
Major zkEVM Projects
- Polygon zkEVM (Type 2) — Mainnet March 2023
- zkSync Era (Type 4) — Mainnet March 2023
- Scroll (Type 1 target) — Mainnet October 2023
- Linea (ConsenSys, Type 3) — Mainnet July 2023
- StarkNet (uses Cairo VM, not EVM) — distinct approach
History
| Year | Event |
|---|---|
| 2019 | First academic papers on ZK proofs for general computation (PLONK, STARK) |
| 2020 | Hermez (later Polygon Hermez) launches ZK token transfer rollup (not full EVM) |
| 2021 | Multiple teams announce zkEVM development; considered “holy grail” difficulty |
| Mar 2023 | Polygon zkEVM and zkSync Era both launch mainnets |
| Jul 2023 | Linea (ConsenSys) mainnet |
| Oct 2023 | Scroll mainnet |
| 2024 | ZK proof times decrease significantly; zkEVM adoption grows across DeFi |
Common Misconceptions
“zkEVM is perfectly compatible with Ethereum.” It depends on the type. Type 1 zkEVMs aim for full compatibility, but most deployed zkEVMs today are Type 2-4 with some incompatibilities. Testing contracts on the specific zkEVM is necessary.
“ZK = anonymous.” Zero-knowledge in zkEVM refers to a mathematical proof technique, not privacy. zkEVM transactions are typically public unless additional privacy is layered on top.
Criticisms
- Proving times are still slow (minutes to hours for final Ethereum settlement)
- Proving requires significant computational resources (specialized hardware)
- Type 4 zkEVMs have breaking differences with Ethereum tooling, limiting ecosystem compatibility
- Multiple competing implementations fragment developer energy
Social Media Sentiment
zkEVM is considered the “endgame” for Ethereum L2 scaling by much of the Ethereum research community. The competitive race between Polygon, zkSync, Scroll, and Linea generates constant technical debate. Ethereum researchers celebrate ZK progress; pragmatic developers question which zkEVM to build on. ZK proofs are viewed as a long-term upgrade to both L2s and Ethereum itself (eventual ZK-based L1 proofs).
Last updated: 2026-04
Related Terms
Sources
Gabizon, A., Williamson, Z. J., & Ciobotaru, O. (2019). PLONK: Permutations over Lagrange-bases for Oecumenical Noninteractive arguments of Knowledge. IACR ePrint.
Ben-Sasson, E., et al. (2018). Scalable, Transparent, and Post-quantum Secure Computational Integrity. IACR ePrint (STARKs).
Buterin, V. (2022). The Different Types of ZK-EVMs. Vitalik.ca.
Polygon Labs. (2023). Polygon zkEVM: A Technical Overview. Polygon Blog.
Matter Labs. (2022). zkSync Era: Scaling Ethereum with Zero-Knowledge Proofs. Matter Labs.