oss-fuzz-gen and fuzz4all
These are competitors: both use LLMs to generate fuzzing inputs and test cases, but OSS-Fuzz-Gen is specifically integrated with Google's OSS-Fuzz infrastructure while Fuzz4All is a language-agnostic fuzzing framework, so users would typically choose one approach based on whether they need OSS-Fuzz integration or broader language support.
About oss-fuzz-gen
google/oss-fuzz-gen
LLM powered fuzzing via OSS-Fuzz.
This framework helps software security teams automate and enhance their fuzz testing efforts by using Large Language Models (LLMs) to generate new fuzz targets for C, C++, Java, and Python projects. It takes existing project code and an LLM as input, then outputs new fuzzing code and detailed reports on its effectiveness, including crash discovery and code coverage. This is intended for security engineers and quality assurance professionals focused on identifying vulnerabilities in open-source and proprietary software.
About fuzz4all
fuzz4all/fuzz4all
🌌️Fuzz4All: Universal Fuzzing with Large Language Models
This tool helps software developers and security researchers automatically find weaknesses in programs that accept various types of code or data as input. You provide the program you want to test and a large language model (LLM), and it generates diverse, realistic inputs to probe for bugs and vulnerabilities. It's designed for those who need to rigorously test the robustness and security of their software.
Related comparisons
Scores updated daily from GitHub, PyPI, and npm data. How scores work