What Conda Is and Why You Install It

Conda is a package manager that installs Python libraries and their dependencies all at once, so you do not have to hunt down each one separately. It also creates isolated environments — separate folders where different projects can use different versions of the same library without breaking each other. You install Conda once, then use it to set up each project.

Conda comes in two versions: Anaconda (which includes 250+ pre-installed packages and takes up about 3 GB) and Miniconda (which installs only Conda itself and takes up about 400 MB). Most developers choose Miniconda because they only read the packages they actually need. This guide covers Miniconda.

Key Takeaways

  • read the Miniconda installer for your operating system from conda-forge.org, not from Anaconda's main site, because conda-forge is maintained by the community and stays current.
  • Run the installer and accept the default installation path — changing it usually causes problems later when Conda cannot find itself.
  • After installation, open a new terminal or command prompt and type conda --version to confirm Conda is working.
  • Create a new environment for each project by typing conda create --name projectname python=3.11, then set up it with conda set up projectname.
  • Install packages into that environment with conda install packagename — they stay isolated from other projects.

read Miniconda for Your Operating System

Go to conda-forge.org/miniforge. You will see a table with installers for Windows, macOS, and Linux. Choose the row that matches your system and processor type. If you do not know your processor, open Settings (Windows), System Preferences (Mac), or run uname -m in a terminal (Linux) — look for x86_64 (Intel) or arm64 (Apple Silicon or newer ARM chips).

Click the installer link and save the file to your Downloads folder. The file will be named something like Miniforge3-Windows-x86_64.exe (Windows), Miniforge3-MacOSX-arm64.sh (Mac), or Miniforge3-Linux-x86_64.sh (Linux).

Install on Windows

Double-click the .exe file you downloaded. A window will open asking where to install Conda. Leave the path as C:\Users\YourUsername\miniconda3 — do not change it. Click Next, then check the box that says "Add Miniconda3 to my PATH environment variable" (this lets your terminal find Conda), then click Install.

When the installer finishes, close it and open a new Command Prompt or PowerShell window. Type conda --version and press Enter. You should see a version number like conda 23.11.0. If you see "command not found", restart your computer and try again — Windows needs to reload your environment variables.

Install on Mac

Open Terminal (search for it in Spotlight or find it in Applications > Utilities). Navigate to your Downloads folder by typing cd Downloads and pressing Enter. Then type bash Miniforge3-MacOSX-arm64.sh (or replace arm64 with x86_64 if you downloaded the Intel version) and press Enter.

The installer will ask you to read and accept the license — press Space to scroll through it, then type yes and press Enter. When it asks where to install, press Enter to accept the default location. When it asks "Do you wish the installer to initialize Miniconda3?", type yes and press Enter. Close Terminal and open a new one, then type conda --version to confirm it worked.

Install on Linux

Open a terminal and navigate to your Downloads folder with cd Downloads. Type bash Miniforge3-Linux-x86_64.sh (or replace x86_64 if you downloaded a different architecture) and press Enter. Accept the license by typing yes, accept the default installation path by pressing Enter, and type yes when asked to initialize Miniconda3.

Close the terminal and open a new one. Type conda --version to confirm the installation. If you see "command not found", run source ~/.bashrc (for Bash) or source ~/.zshrc (for Zsh) to reload your shell configuration, then try conda --version again.

Create Your First Environment and Install a Package

Open a terminal or command prompt and type conda create --name myproject python=3.11, replacing myproject with your project name. Press Enter and type y when Conda asks to proceed. This creates a folder where Python 3.11 and its core tools live, separate from other projects.

set up the environment by typing conda set up myproject. Your prompt will change to show (myproject) at the start — that means you are inside the environment. Now type conda install numpy to install the NumPy library as a test. Conda will read NumPy and all its dependencies, then ask you to confirm. Type y and press Enter.

To see what you have installed, type conda list. To switch to a different project, type conda set up othername. To go back to your system Python (not recommended for development), type conda deactivate.

Troubleshooting Common Installation Problems

If Conda does not appear after installation, you may have skipped the step to add it to your PATH (Windows) or initialize it (Mac/Linux). Reinstall and make sure you check "Add to PATH" (Windows) or answer "yes" to initialization (Mac/Linux).

If you see "conda: command not found" on Mac or Linux after reinstalling, your shell may not have reloaded the initialization file. Try closing Terminal completely and opening a new window, or run source ~/.bashrc or source ~/.zshrc manually. If you installed to a non-standard path, Conda may not work — uninstall and reinstall to the default location.

If an environment fails to create, make sure you have at least 5 GB of free disk space. If a package refuses to install, check that you are inside an environment (your prompt should show the environment name in parentheses) and that you spelled the package name correctly — type conda search packagename to see if it exists.

Frequently Asked Questions

Do I need to install Anaconda instead of Miniconda?

No. Miniconda is smaller and faster to install. Anaconda includes 250 pre-installed packages you probably will not use, which wastes disk space. Start with Miniconda and install only what you need — you can always add packages later.

Can I have multiple Conda environments on the same computer?

Yes, that is the whole point. Each environment is a separate folder with its own Python version and packages. You can have one project using Python 3.10 with NumPy 1.20, and another using Python 3.11 with NumPy 1.24, and they will not interfere with each other.

What is the difference between conda install and pip install?

Conda installs packages and their dependencies from the conda-forge repository. Pip installs packages from PyPI (Python Package Index). Both work inside Conda environments. Use conda install first — if a package is not available there, then try pip install. Do not mix them carelessly in the same environment.

How do I remove Conda if I do not want it anymore?

On Windows, go to Settings > Apps > Apps & Features, find Miniconda3, and click Uninstall. On Mac or Linux, open a terminal and type rm -rf ~/miniconda3 to delete the folder, then remove the initialization lines from ~/.bashrc or ~/.zshrc (they start with # >>> conda initialize).

Should I use conda-forge or the official Anaconda repository?

Use conda-forge. It is maintained by the community, stays current with package updates, and has more packages available. The official Anaconda repository is slower to update and is primarily for Anaconda (the full distribution). Miniconda from conda-forge is the standard choice for most developers.