Skip to main content

​
System Requirements

DVC requires:
  • Python: 3.9 or higher (3.9, 3.10, 3.11, 3.12, 3.13, 3.14)
  • Git: Installed and configured (unless using --no-scm mode)
  • Operating System: Linux, macOS, or Windows
DVC works best in Git repositories, but you can use it standalone with the --no-scm option during initialization.

​
Quick Install

Choose your preferred installation method:
Install DVC using Python’s package manager:
pip install dvc
Verify the installation:
dvc version
Use a virtual environment to avoid conflicts with other Python packages:
python -m venv dvc-env
source dvc-env/bin/activate  # On Windows: dvc-env\Scripts\activate
pip install dvc

​
Installation with Remote Storage Support

DVC supports multiple remote storage types. Install additional dependencies based on your storage needs:

​
Using pip

# AWS S3
pip install 'dvc[s3]'

# Azure Blob Storage
pip install 'dvc[azure]'

# Google Cloud Storage
pip install 'dvc[gs]'

# Google Drive
pip install 'dvc[gdrive]'

# SSH/SFTP
pip install 'dvc[ssh]'

# Alibaba Cloud OSS
pip install 'dvc[oss]'

# WebDAV
pip install 'dvc[webdav]'

# HDFS
pip install 'dvc[hdfs]'

# WebHDFS
pip install 'dvc[webhdfs]'

# Install all remote storage types
pip install 'dvc[all]'
You can install multiple storage backends at once:
pip install 'dvc[s3,gs,azure]'

​
Using conda

# AWS S3
mamba install -c conda-forge dvc-s3

# Azure Blob Storage  
mamba install -c conda-forge dvc-azure

# Google Cloud Storage
mamba install -c conda-forge dvc-gs

# Google Drive
mamba install -c conda-forge dvc-gdrive

# SSH/SFTP
mamba install -c conda-forge dvc-ssh

# Alibaba Cloud OSS
mamba install -c conda-forge dvc-oss
When using package managers like Snap, Homebrew, or Chocolatey, storage dependencies are included by default.

​
OS-Specific Packages

DVC provides self-contained packages for various operating systems.

​
Ubuntu / Debian (deb)

# Add DVC repository
sudo wget https://dvc.org/deb/dvc.list -O /etc/apt/sources.list.d/dvc.list
wget -qO - https://dvc.org/deb/iterative.asc | sudo apt-key add -

# Update and install
sudo apt update
sudo apt install dvc

​
Fedora / CentOS / RHEL (rpm)

# Add DVC repository
sudo wget https://dvc.org/rpm/dvc.repo -O /etc/yum.repos.d/dvc.repo
sudo rpm --import https://dvc.org/rpm/iterative.asc

# Update and install
sudo yum update
sudo yum install dvc

​
Standalone Packages

Download standalone executables for Linux, macOS, and Windows from the GitHub Releases page.
Standalone packages are self-contained but may be larger in size. They’re useful for environments where you can’t use package managers.

​
Development Version

Install the latest development version from GitHub: 
pip install "dvc @ git+https://github.com/treeverse/dvc"
The development version may contain unreleased features and bugs. Use it only for testing or contributing to DVC.

​
Shell Completion

Enable tab completion for your shell: 
# Add to ~/.bashrc
eval "$(dvc completion -s bash)"
After adding the completion script, restart your shell or source the configuration file.

​
VS Code Extension

Install the DVC Extension for VS Code for a visual interface:
1

Install the Extension

Search for “DVC” in the VS Code Extensions marketplace or install from the command line:
code --install-extension Iterative.dvc
2

Install DVC CLI

The extension requires the DVC command-line tool to be installed on your system. Follow the installation steps above if you haven’t already.
3

Open a DVC Project

Open a folder containing a DVC project (with .dvc directory) to activate the extension.
The VS Code extension provides experiment tracking, data management, plots visualization, and more — all from within your editor.

​
Verifying Installation

Confirm DVC is installed correctly: 
# Check version
dvc version

# View help
dvc --help

# Check which storage backends are available
pip list | grep dvc
You should see output similar to: 
dvc               3.x.x
dvc-data          x.x.x
dvc-objects       x.x.x
dvc-render        x.x.x
# Plus any storage backends you installed
dvc-s3            x.x.x

​
Troubleshooting

If dvc isn’t recognized:
  1. Check your PATH: Ensure the installation directory is in your system’s PATH
  2. Restart your terminal: Close and reopen your terminal
  3. Virtual environment: If using venv, make sure it’s activated
  4. Reinstall: Try reinstalling DVC using a different method
On Linux/macOS, you may need to use sudo for system-wide installations:
sudo pip install dvc
Or install for your user only:
pip install --user dvc
DVC requires Python 3.9+. Check your version:
python --version
If you have multiple Python versions, use pip3 or specify the version:
python3.11 -m pip install dvc
If you get errors like “S3 remote is not supported”:
  1. Install the appropriate storage backend: 
    pip install 'dvc[s3]'
  2. Verify installation: 
    pip list | grep dvc

​
Updating DVC

Keep DVC up to date to get the latest features and bug fixes: 
pip install --upgrade dvc

​
Uninstalling DVC

To remove DVC from your system: 
pip uninstall dvc
Uninstalling DVC doesn’t remove your .dvc directories or cached data. To completely clean up, manually delete .dvc/cache directories.

​
Next Steps

Now that DVC is installed:

Quick Start

Follow our hands-on tutorial to start using DVC.

Initialize a Project

Learn how to set up DVC in your repository.

Configure Remote Storage

Set up remote storage for your data.

Core Concepts

Understand how DVC works under the hood.

Build docs developers (and LLMs) love

Get started for freeTalk to us