Original author: @thirdweb
Original translation: @BitalkNews
zkEVM is bringing Ethereum to the masses. But with so many choices, which one is the best? We compare Polygon zkEVM (+2.0), zkSync Era, Linea, Scroll, and Taiko to bring you the ultimate guide to zkEVM.
A quick recap:
zkEVM is a layer 2 network that handles transactions and sends proofs back to Ethereum. They use zero-knowledge proofs (ZK) and EVM compatibility to reduce computational load on Ethereum (L1) and improve scalability without sacrificing security or decentralization.
However, not all zkEVMs are the same, there are different types of zkEVM:
Vitalik has a great framework that showcases the pros and cons of each type, from Ethereum equivalent (Type 1) to EVM compatible (Type 4). The main tradeoff is EVM compatibility versus performance:
So, which zkEVM is the best? We compare a few popular options:
→ Polygon zkEVM (+2.0)
→ zkSync Era
→ Linea
→ Scroll
→ Taiko
Let's break them down.
Polygon 2.0 (zkEVM Validium)
In June 2023, Polygon Labs proposed to upgrade Polygon PoS to zkEVM validium.
Polygon PoS has a strong ecosystem:
→ 2B+ TVL (Total Value Locked)
→ Thousands of dApps
→ About 2.5 million transactions per day
Migrating PoS to zkEVM validium will preserve its network effects and low fees without impacting its existing applications, users, and developer experience. But wait, doesn't Polygon already have zkEVM?
Polygon believes that 2.0 (the proposed zkEVM validium) is a lower-cost, higher-throughput version of its Polygon zkEVM Rollup.
Rollups use Ethereum to publish transaction data and verify proofs, inheriting its security and decentralization. However, there are trade-offs:
The drawback of Rollup is the high cost and limited throughput of publishing transaction data to Ethereum. Validium provides similar security guarantees as Rollup (ZK proofs ensure transaction validity), but transaction data is provided off-chain.
Compared to Rollup, this gives Validium two main advantages:
Lower cost, as they do not consume expensive Ethereum Gas (block space for storing transaction data)
Higher scalability, as Rollups' throughput is limited by the transaction data that can be published to Ethereum
The trade-off of Validium is that it must ensure the availability of transaction data outside of Ethereum, which can be challenging. However, Polygon PoS's existing 100+ validators serve as highly secure and reliable guarantees for data availability.
So why use Polygon zkEVM?
Polygon zkEVM
Polygon zkEVM is a type 3 ZK-Rollup. Its goal is to be functionally equivalent to the EVM - meaning it has greater scalability when working with existing Ethereum tools and clients (by preserving EVM opcode compatibility).
Currently, Polygon PoS (2.0) and Polygon zkEVM rollup are two public networks in the Polygon ecosystem. After the upgrade, this remains the case: one as Rollup, the other as Validium.
But why not just abandon Polygon zkEVM directly?
Polygon 2.0 and Polygon zkEVM complement each other: Polygon zkEVM (Rollup) has slightly higher costs and lower throughput but offers the highest level of security. This is best suited for applications handling high-value transactions (such as DeFi).
On the other hand, Polygon 2.0 (zkEVM validium) will provide the highest scalability. This is best suited for applications with high transaction volume and low cost (such as Web3 games, social platforms).
Polygon zkEVM's mainnet will launch in March 2023, with a current TVL of approximately $54 million on the network. Built on Polygon zkEVM: https://thirdweb.com/polygon-zkevm
zkSync Era
zkSync Era is currently a type 4 zkEVM, but over time, it may increase compatibility with EVM bytecode. Era (its mainnet) is built by Matter Labs and went live in March 2023, with a current TVL of approximately $450 million and a transaction volume of 27 million in the past month.
zkSync's goal is not EVM equivalence but improving proof generation speed through its custom virtual machine (VM). It natively supports Ethereum cryptographic primitives, simplifies testing for Hardhat plugins, and improves user experience through native account abstractions.
zkSync Era uses its LLVM compiler to convert Solidity, Vyper, and Yul code into executable code on the zkSync VM and plans to be compatible with Rust and C++ in the future. It has a powerful L1-L2 contract message passing system that helps developers transfer data between Ethereum and zkSync contracts.
The upcoming zkPorter integration in zkSync will provide users with the following choices:
- High-security zkRollup accounts with approximately 20x lower fees compared to Ethereum
- zkPorter accounts with stable transaction fees and different security models, allowing for flexible user experiences
For data availability, zkSync Era focuses on state diffs instead of transaction inputs—providing data compression and integration with zkPorter.
As the only type 4 zkEVM in this list, its custom virtual machine enables certain features (along with higher scalability) on zkSync Era.
For example, zkSync Era enhances developer experience by natively supporting account abstractions using its custom virtual machine. However, EVM-equivalent chains cannot natively support account abstractions and achieve them through Ethereum's ERC-4337.
Type 4 zkEVM's drawback is lack of compatibility - development process needs to be adjusted to build applications on zkSync. Build on zkSync Era: https://twitter.com/thirdweb/status/1666119985520066561
Linea
Built by ConsenSys (creators of MetaMask, Truffle, and Infura), Linea is currently a type 3 zkEVM. It integrates with native MetaMask wallets and the Truffle development toolkit, and its Alpha mainnet was launched in July 2023 with a TVL of $50 million.
Linea's uniqueness lies in its ability to execute unaltered native bytecode based on the Ethereum Virtual Machine (EVM) specification. Its goal is to provide a type 2 zkEVM in the future, ensuring near-full compatibility with Ethereum while maintaining scalability.
Linea does not use a transpiler or custom compiler. It directly uses compiled Solidity bytecode to generate ZK proofs for smart contracts, reducing the risk of errors and hacker attacks. For users, this means similar security to Ethereum but with much lower transaction costs.
Build on Linea: https://thirdweb.com/linea
Scroll zkEVM
Scroll is a zk-Rollup on Ethereum and is currently a type 3 zkEVM with plans to transition to type 2.
Its long-term goal is to prioritize EVM compatibility and decentralization over fast ZKP generation, ultimately shifting towards type 1 zkEVM.
A successful type 1 zkEVM is the ultimate form of EVM compatibility, decentralization, and scalability - essentially replicating Ethereum L1. However, as we have seen with Ethereum itself, achieving this is much more difficult.
Currently, Scroll's alpha testnet is a type 3 zkEVM and is open to the public. Build on Scroll: https://thirdweb.com/scroll-alpha-testnet
Taiko
Similar to Scroll, Taiko is a general-purpose zk-Rollup aiming to be a type 1 native zkEVM. Taiko takes a different approach than Scroll to become type 1, prioritizing decentralization from the start rather than speed:
Scroll is gradually enhancing EVM compatibility (starting from the Type 3 testnet) and making its sorter and prover more decentralized at mainnet launch. Despite slower ZKP generation speed, Taiko decentralized these components from the start and maximized EVM compatibility.
Taiko's Type 1 testnet is currently available. Build on Taiko: https://thirdweb.com/taiko-grimsvotn-l2
Using Vitalik's framework: → Taiko: Type 1 (testnet live) → Polygon zkEVM: Type 3 (mainnet live + planned migration to Type 2) → Linea: Type 3 (mainnet live + planned migration to Type 2) → Scroll: Type 3 (testnet live + planned transition to Type 1) → zkSync Era: Type 4 (mainnet live)
In summary, L2 development and adoption are crucial for Ethereum's future + solving blockchain scalability issues. Each zkEVM has unique advantages and faces unique challenges, but they all play a critical role in achieving an accessible web3 ecosystem.
