Empirical Study of Zero-Knowledge Proof (ZKP) Languages

This repository presents an empirical study of Zero-Knowledge Proof (ZKP) languages (including zkVMs), offering an overview of the most prominent ones. The study leverages the GitHub API to gather and analyze key metrics about repositories related to each language (e.g., number of stars, date of last update, etc.).

Overview

This repository includes:

• A quantitative analysis of repositories for many ZKP languages using metrics such as:
  • Number of stars
  • Number of open issues
  • Date of last update
  • Number of contributors
• Visualizations to highlight interesting trends.
• All code used for data collection and visualization.

Studied ZKP Languages

This study covers the following ZKP languages:

Name	Type
Circom	DSL
ZoKrates	DSL
Noir	DSL
Cairo	DSL
Leo	DSL
Nexus VM	zkVM
RISC Zero	zkVM
snarkVM	zkVM
SP1	zkVM
Jolt	zkVM

• DSL: Domain-Specific Language, designed specifically for writing ZKP programs.
• zkVM: Zero-Knowledge Virtual Machine, a virtual machine designed to execute programs written in standard programming languages while generating zero-knowledge proofs of their correctness.

Visualizations

Figures are provided to illustrate:

• Number of repositories with more than 1 star for each language (with DSL/zkVM distinction).
• Combined repository activity chart showing total repositories with recently updated ones highlighted.
• Number of repositories updated after 1 January 2024 for each language (with DSL/zkVM distinction).
• Number of repositories with more than 10 total issues for each language (with DSL/zkVM distinction).
• Percentage distribution of all DSL repositories.
• Percentage distribution of all zkVM repositories.
• Comparison of repository counts: Circom vs zkVMs.

GitHub API Search Queries

The study utilized GitHub's API to search for repositories. Below are the specific search queries used to identify relevant repositories for each language:

Language	GitHub Search Query
Circom	filename:.circom
ZoKrates	filename:.zok
Noir	filename:nargo.toml
Cairo	filename:scarb.toml
Leo	filename:.leo
Nexus VM	nexus_rt
RISC Zero	risc0-build
snarkVM	snarkvm
SP1	SP1_zkvm
Jolt	jolt-sdk

Explanation of Queries:

• File-based Searches (filename:):
  Some queries, like filename:.circom for Circom and filename:scarb.toml for Cairo, target specific file extensions or package manager files unique to the language.
• Build Tools or SDKs:
  Queries such as risc0-build and nexus_rt focus on build tools or SDKs that are central to these zkVMs.
• zkVM Identifiers:
  For zkVMs like SP1 and SNARKVM, the queries target unique identifiers (SP1_zkvm and snarkvm) associated with these projects.

Results

The analysis revealed the following key insights:

• Circom appears to be the most popular ZKP language, likely due to its strong performance and optimization capabilities. Cairo is also gaining significant traction, which aligns with its role as the core language of StarkNet.
• RISC Zero stands out as the most widely adopted zkVM, most likely because it is formally verified and contains strong developer tooling.
• zkVMs have seen great usage from 2024, surpassing Circom in growth trends. This shift is mainly due to their usability. Developers no longer need to manually write circuits; instead, they can write "standard code" and rely on the underlying VM to generate the necessary proofs. However, when performance is critical, Circom remains the preferred choice due to its low-level control.
• ZoKrates shows very limited usage, likely due to its relatively lower performance and stagnated development (with the last commit recorded in April 2024).

Repository Structure

• figures/
  Contains visualizations of the study in .png format.

• metrics/
  Contains .csv files with detailed metrics for each language.

• src/
  Contains source files for data collection and analysis:

  • data_analyzer.ipynb: Jupyter notebook for analyzing and visualizing data.
  • data_collector.py: Python script for fetching data using the GitHub API.
  • requirements.txt: Lists Python dependencies required for running the project.

• README.md
  This file.

Getting Started

Prerequisites

• Python 3.8 or later
• Git
• Jupyter Notebook (for running .ipynb files)

Installation

1. Clone the repository:

   git clone https://github.com/ArmanKolozyan/ZKP-Languages.git
   cd zkp-languages

2. Install dependencies:

   cd src
   pip install -r requirements.txt

Usage

1. Fill in your GitHub API token:

   • Generate a token at GitHub Personal Access Tokens.
   • Add your token in the appropriate location in data_collector.py.

2. Run data_collector.py to collect data. When prompted, enter the desired search term in the terminal:

   python src/data_collector.py

3. Run all cells in data_analyzer.ipynb to process and visualize the data:

   jupyter notebook src/data_analyzer.ipynb

Contributing

Contributions are welcome! Please create an issue or submit a pull request if you have ideas to improve the project.

TODO

• Support more ZKP languages
  • Noir
  • Leo
  • Jolt zkVM
  • Zinc
  • Halo2
• Add more details to the languages table.
• Manually analyze GitHub issues of the top 5 repositories for each language.
• Manually clean the .csv files.
• Add more conclusions to the results.
• Conduct a historical analysis: e.g., comparing number of zkVM repositories to Circom repositories over the years.
• Distinguish between the 2 use cases of Cairo (writing smart contracts vs. proof-related applications).
• Explore GitHub’s dependency tracking feature (e.g., https://github.com/risc0/risc0/network/dependents).
• Design a front-end dashboard for improved data visualization.