Skip to main content
Environments in conda are self-contained, isolated spaces where you can install specific versions of software , including dependencies, libraries, and Python versions. This isolation helps avoid conflicts between package versions and ensures that your projects have the exact libraries and tools they need. For more information on environments, see our environments concepts page.

​
Working with environments

For convenience, the most common actions users take when managing environments are detailed here. For a full list of actions and more comprehensive guide, see Manage environments in the official conda documentation. Alternatively, follow along with our Getting started with conda environments tutorial on Anaconda.com.
If you prefer to create and manage your environments via our graphical interface, Navigator, see Managing environments.

​
Creating an environment

Create a by opening Anaconda Prompt (Terminal on macOS/Linux) and following the steps for your use case below:
The following command creates a new environment and downloads the specified packages and their dependencies:
conda create --name <ENV_NAME> <PACKAGE>=<VERSION> <PACKAGE> <PACKAGE>
Replace <ENV_NAME> with a name for your environment.
Replace each <PACKAGE> with the name of your desired packages.
Replace <VERSION> with your desired version (optional).
Example environment creation command
conda create --name myenv python=3.11 beautifulsoup4 docutils jinja2=3.1.4 wheel
Version matching utilizes MatchSpec Protocol. Below are some examples for specifying versions for packages when creating an environment. If you do not specify a package version, conda will attempt to install the latest version of the package from its available .
  • python=3.12.1 - Matches the package with the exact name and version specified.
  • python>=3.11 - Matches any version of the package that is greater than or equal to the version specified.
  • python<=3.12 - Matches any version of the package that is less than or equal to the version specified.
  • python>3.10,<3.12 - Matches any version of the package between the specified versions, but excludes the specified versions.
  • python>=3.10,<=3.12 - Matches any version of the package between the specified versions, including the specified versions.
  • python[version='3.12.*'] - Matches any version of the python package that starts with 3.12.
It is best practice to install all of the packages that you need in your environment in a single command. This reduces the risk of dependency conflicts or broken environments.
When you create an environment, conda it to the environments.txt file, which is located here:
%USERPROFILE%\.conda\environments.txt
conda --info envs uses the environments.txt file to show all existing environments on your machine, even those outside the base install directory.

​
Environment locations

By default, conda creates environments in the following locations, depending on your operating system and how you’ve installed conda:
Operating SystemInstallerDefault Environment Location
WindowsGraphical installer (.exe)
  • C:\Users\<USERNAME>\anaconda3\envs\
  • C:\Users\<USERNAME>\miniconda3\envs\
macOSGraphical installer (.pkg)
  • /opt/anaconda3/envs/
  • /opt/miniconda3/envs/
macOSCommand line installer (.sh)
  • /anaconda3/envs/
  • /miniconda3/envs/
LinuxCommand line installer (.sh)
  • /anaconda3/envs/
  • /miniconda3/envs/
You can specify the path where you want the environment to be created by including the --prefix command when creating an environment, or by updating your .condarc file to contain the envs_dirs: key.
Example .condarc configuration
channels:
  - defaults
default_channels:
  - https://repo.anaconda.com/pkgs/main
envs_dirs:
  - /Users/<USER>/conda-envs

​
Activating an environment

Because environments are isolated spaces, you can only work with one at a time. Selecting an environment to work with is called activating it. Activate an environment by running the following command: 
conda activate <ENV_NAME>
Replace <ENV_NAME> with the name of the environment you want to activate.

​
Switching between environments

When you’re ready to switch between projects, simply activate the environment of your other project. Activating a different environment will deactivate your current one.
  1. (Optional) View a list of all your environments by running the following command: 
    conda info --envs
  2. To switch to a different environment, activate it by running the following command: 
    conda activate <ENV_NAME>
    Replace <ENV_NAME> with the name of the environment you want to switch to.

​
Locking an environment

The most reliable way to ensure your project remains reproducible indefinitely is to “lock” its environment. Locking an environment creates a fully specified environment, one that has all packages used in the project and their dependencies configured to a specific version. This ensures that your project will be reproduced exactly as it was initially configured, because there will never be an unexpected update or change if new package dependencies are released. Locking your project requires the conda-project package to be installed in the environment you want to lock. Install the package by running the following commands: 
conda activate <ENV>
conda install conda-project
Replace <ENV> with the environment you want to lock.
If your project doesn’t contain an environment.yml file, create one by running the following command: 
conda-project init
You can then lock your project’s environment by running the following command: 
conda-project lock
Locking your project produces a conda-lock.default.yml file that you can export to share with others.

​
Sharing an environment

Sharing your environment with someone else allows them to use conda to recreate your environment on their machine. To share an environment and its software packages, you must export your environment’s configurations into a .yml file.
Simply copying your Anaconda or Miniconda files over to a new directory or another machine will not recreate the environment. You must export the environment as a whole.

​
Exporting the environment configuration .yml file

If you already have an environment configuration .yml file in your current directory, it will be overwritten during the export process.
  1. Activate the environment you want to export by running the following command: 
    conda activate <ENV_NAME>
    Replace <ENV_NAME> with the name of the environment you want exported.
  2. Export the environment by running the following command: 
    conda env export > environment.yml
    The environment.yml file populates in your current working directory.
    This operation includes both the environment’s conda and pip packages.
  3. Share the exported environment configuration .yml file with another user.

​
Deactivating an environment

It is best practice to deactivate your environment when you are finished working in it. To deactivate your active environment, run the following command: 
conda deactivate
When you deactivate an environment, conda returns to the previously activated environment.For example, if you run conda activate my_env1, followed by conda activate my_env2, and then run conda deactivate, conda returns to the my_env1 environment.

​
Removing an environment

To remove an environment, run the following command: 
conda remove --name <ENV_NAME> --all
Running conda remove deletes and your environment.
If, for any reason, you need to delete an environment directory manually, don’t use your file explorer. If you do, you’ll remove the environment’s contents and free up space on your machine, but leave the environment path listed in your environments.txt file (see Environment registration).To manually delete and deregister an environment directory properly, run one of the following commands, depending on your operating system and conda installation:
C:\Users\<YOUR_USERNAME>\anaconda3\_conda constructor uninstall --prefix <PATH_TO_ENV_DIRECTORY>
Replace <YOUR_USERNAME> with your Windows username.
Replace <PATH_TO_ENV_DIRECTORY> with the path to the directory that contains the environments.

​
Troubleshooting

​
command not found: conda

Cause
The command not found: conda error occurs when your command line interface (CLI) can’t find the conda command in order to use it. This might be because:
  • You don’t have conda properly initialized.
  • You have set auto_activate_base to false.
  • You’re using a shell that conda doesn’t support.
  • Conda is not installed or the install was incomplete or corrupted.
These issues primarily occur on macOS/Linux computers. Anaconda Distribution and Miniconda installations on Windows include Anaconda Prompt, which opens with conda initialized by default.
Solution
If you recently installed Anaconda Distribution or Miniconda, make sure you closed and reopened your CLI to make conda’s initialization take effect.
If you don’t want to close your CLI, you can also use one of the following source commands to refresh your shell:
source ~/.bashrc
Use the command that matches your shell.
You can also initialize conda directly from its bin directory:
<PATH_TO_CONDA>/bin/conda init
Replace <PATH_TO_CONDA> with a path to your conda installation.
To see the value for auto_activate_base, run the following command:
conda config --describe auto_activate_base
If your terminal returns false, this means that conda is not automatically activating your base environment when you start a new shell. This behavior emulates your system Python, and some users prefer to have their conda environment be inactive until they need it. However, this is not conda’s default behavior after installation.To change the value of auto_activate_base to true, run the following command:
conda config --set auto_activate_base true
If you have auto_activate_base set as false, the conda command will still be available as a shell function, but your base environment will not be active when a new shell is started. To activate your base environment, run conda activate.
For information on which shells conda supports, see Conda activate in the official conda documentation.
If you have tried to initialize conda in your shell but it didn’t work, try uninstalling and reinstalling Anaconda Distribution or Miniconda.
Make sure that you say yes to the initialization option, and, if installing from the CLI, reinitialize your shell by restarting it or using its source command. For more information, see Initialize conda in your shell.

​
DirectoryNotACondaEnvironmentError: The target directory exists, but it is not a conda environment

Cause
This error means that you don’t have a conda environment currently active. You might have used conda deactivate while in your base environment and deactivated conda.
Solution
Activate your base environment or another environment:
# Activate your base environment
conda activate

# Activate an environment by name
conda activate <ENV_NAME>
Replace <ENV_NAME> with the name of the environment.

​
PackagesNotFoundError: The following packages are not available from current channels

Cause
The PackagesNotFoundError occurs when conda cannot find the requested packages in any of your currently configured channels in order to install or update it. This might be because:
  • The package name is misspelled.
  • The package doesn’t exist in the channels included in your conda configuration file (.condarc).
  • The package is not available for your operating system or architecture.
Solution
Verify that your package name is spelled correctly and that your command contains no syntax errors. For more information about command syntax, see Using conda install.
If you’re unsure what the correct command to install the package is, you can find it on the package’s details page.Search Anaconda.org or your Anaconda Platform channels for your desired package, and view its details. Each package’s details page provides a copyable command to install the package from that channel. Run that command to install the package in your currently active environment.
Remember to pay close attention to the command’s syntax! The command you copy might not be the exact command you want to run, depending on your use case. Anaconda.org uses the double-colon syntax for channel specification, while Anaconda Platform uses the channel flag syntax by default. For more information, see Installing packages from a specific channel.
Add the missing channel to your .condarc file. However, keep in mind that any channels listed in the channels: section of your .condarc file will be searched by conda every time you install a new package.
Check your system architecture to determine if the package is compatible. This information is found under platform: in your conda info output.
conda info
Conda restricts searches to packages that are compatible with your system architecture. If the package you’re looking for is not available for your system architecture, it is not listed in conda’s package search results.