This page explains how to set up your environment for using FuzzBench.

Getting the code

Clone the FuzzBench repository to your machine by running the following command:

git clone
cd fuzzbench

Installing prerequisites


Install Docker using the instructions here. Googlers can visit go/installdocker.

If you want to run docker without sudo, you can create a docker group.

To ensure there are no problems building docker images, we recommend using a recent docker release. If you are using an older release, and your builds are failing you may need to enable BuildKit. This step is unnecessary on recent Docker releases.

Note: Docker images can consume significant disk space. Clean up unused docker images periodically. You can do this with docker-cleanup to garbage collect unused images.


Install make for your linux distribution. E.g. for Ubuntu:

sudo apt-get install build-essential

Python programming language

Download Python 3.10, then install it.

If you already have Python installed, you can verify its version by running python3 --version. The minimum required version is 3.10.8.

Python package dependencies

Install the python dependencies by running the following command:

sudo apt-get install python3.10-dev python3.10-venv
make install-dependencies

This installs all the dependencies in a virtualenv .venv. Activate this virtualenv before running further commands.

source .venv/bin/activate

You can exit from this virtualenv anytime using the deactivate command.


You can verify that your local setup is working correctly by running the presubmit checks.

make presubmit


You can format your changes using the following command:

make format

Local Support

If you want to run FuzzBench locally on your own machines or servers, they need rsync installed:

sudo apt-get install rsync