Kakarot: A Zk-EVM that Vitalik Buterin thinks is cool and has invested in.

Written by: Yinan


Kakarot zkEVM is an Ethereum Virtual Machine (EVM) written in Cairo. Cairo is a Turing-complete language associated with the CairoVM. CairoVM implements provable computation using polynomials and the ZK-STARK proof system.

The distinguishing feature of zkEVM is its ability to generate provable transactions, and therefore, provable blocks. Kakarot is built on top of CairoVM, so every transaction executed on Kakarot is provable.

Overall, Kakarot zkEVM is written in the Cairo language and uses the ZK-STARK proof system to provide provable computation. It enables developers to deploy EVM applications on top of it and allows end-users to interact with those applications using their familiar toolchains.

Finally, Kakarot will also provide interoperability with the Starknet protocol, enabling composability across different domains such as DeFi and GameFi protocols.


The project started in October 2022 with a few objectives: open-sourced from day one, and the most ambitious goal was to create a zkEVM with less than 10,000 lines of code. This goal was achieved in just six months. And a testnet will be released by the end of summer 2023.

Kakarot zkEVM’s roadmap is divided into three phases:

  • Phase One: Kakarot zkEVM on Starknet — Bringing EVM to Starknet

    Kakarot zkEVM will be initially integrated as a built-in EVM inside Starknet’s L2, enabling developers to deploy their Solidity (or any EVM-compatible language) smart contracts directly on Starknet and use their familiar tools (such as Foundry, Hardhat, Wagmi, etc.). Their end-users will be able to interact with their dApps using their familiar toolchains (such as Metamask, Wallet connect, etc.). Overall, developers and users will have the same experience on Kakarot as they would have on Polygon, Scroll, or Ethereum L1.

  • Phase Two: Kakarot x Madara — L3 zkEVMs

    Kakarot and Madara will be integrated into a unified stack to enable L3 zkEVMs and enable L4, L5, etc where meaningful. Teams will be able to deploy their zkEVM application chains and settle transactions on Starknet using validity proofs. Computation is done off-chain or at higher layers, and verification is done on-chain. These L3s that rely on validity proofs (such as Kakarot) have an interesting and underappreciated property: the ability to decouple security and decentralization. Users can benefit from L1 security without the need for the same level of decentralization (i.e., thousands of validators). However, decentralization is beneficial for rollups because it brings liveness and censorship resistance, which are two very important properties. This can be achieved by setting the sequencer at hundreds instead of thousands. Gas costs on L2 are much cheaper than L1, and performance (TPS) is higher because computation is done off-chain. Note that L2 is already much cheaper than L1. The scalability stack of rollups compounds. To further reduce gas costs, proof verification and data availability (DA) can be separated. The Starknet L2 can serve only as a proof verification layer, and new data availability solutions such as Celestia or EigenDA can be used to publish transaction data. Users will have the choice to opt for either option depending on their security needs. Publishing proofs and transaction data on Starknet is the more secure option, while publishing transaction data using a DA solution is the more economical choice.

  • Phase Three: Kakarot x Madara — type 1 zkEVM

    Kakarot and Madara can also be combined to achieve a type 1 zkEVM if:

    1. The Ethereum consensus rules can be written in Cairo inside a Madara x Kakarot full node, enabling L1 consensus to be proven.
    2. A switch from Pedersen-Merkle-Blockingtricia Trie (MPT) to Keccak MPT is made.

    Then, Kakarot becomes a type 1 zkEVM client, capable of proving L1 blocks. This is a higher-level use case depending on Ethereum’s roadmap, particularly Verge. After Verge, Keccak may be replaced by Poseidon as Ethereum’s preferred hash function. This will help the zkEVM team become type 1 because the main compatibility blocker for zkEVM is the storage layout, i.e., implementing Keccak MPT with provable and reasonable costs.


After a week of focused work in a hacker house in Lisbon, about ten people built a zkEVM, and thus the core team of Kakarot was born. With the help of Starknet Foundation and Onlydust, Kakarot now has over 40 unique contributors.

Elias Tazartes: Co-founder and CEO of Kakarot, a freelance developer specializing in blockchain engineering (Dapp, Ethereum, Algorand) and full-stack development (Svelte/React, NodeJS, Rust, Python).

Clément Walter: Scientist, engineer, entrepreneur, and handyman, curious about everything. This is Clément’s self-introduction. He is also a co-founder and CTO of Kakarot.


Currently, Kakarot has conducted a pre-seed round of financing, and the specific amount of financing has not been announced. The teams and individuals participating in the investment include Starkware, LambDa. Angel investors include Vitalik, co-founder of Ledger Nicolas Bacca and co-founder and CEO of ZAMA Rand Hindi.


Currently, the development of public chain ecosystems that are not compatible with EVM is slow. Most of the public chain ecosystems revolve around the Ethereum ecosystem. Kakarot, as an Ethereum virtual machine (EVM) written in Cairo on Starknet, can well link the Starknet ecosystem and the Ethereum ecosystem, allowing the Ethereum ecosystem to be well built on Starknet, and reduce the entry barriers for developers, while accessing Ethereum security and reducing costs. In the second and third stages of Kakarot, more composability can be achieved, such as modular combinations such as Celestia or EigenDA.

As the ZK-EVM written on Starknet, the connection with the Starknet ecosystem is like a relationship between lips and teeth, flourishing together. Currently, Starknet is also constantly iterating and upgrading to improve user experience, and the ZK-EVM built on Kakarot can contribute well to the ecological prosperity of the project. Conversely, if Starknet’s development is not satisfactory, Kakarot will also be affected.

Another advantage of Kakarot is that it has received support from Vitalik and StarkWare, not only spiritually but also financially.

However, compared with other ZK-EVMs, Kakarot is at a disadvantage – a team of less than 10 people, with a much smaller budget and brand. This is also a reason why Kakarot has not been deployed as a universal zkEVM L2.

Kakarot is currently in its early stages, and the testnet will be released soon. Please stay tuned.

Like what you're reading? Subscribe to our top stories.

We will continue to update Gambling Chain; if you have any questions or suggestions, please contact us!

Follow us on Twitter, Facebook, YouTube, and TikTok.


Was this article helpful?

93 out of 132 found this helpful

Gambling Chain Logo
Digital Asset Investment
Real world, Metaverse and Network.
Build Daos that bring Decentralized finance to more and more persons Who love Web3.
Website and other Media Daos

Products used

GC Wallet

Send targeted currencies to the right people at the right time.