conda env export

Export a conda environment to a file or stdout

Overview

The conda env export command exports a given environment’s package specifications to a file or stdout. This is essential for creating reproducible environments and sharing environment specifications with others.

This command is also available as conda export (without the env subcommand).

Syntax

conda env export [options]

By default, exports the currently active environment to stdout in YAML format.

Target Environment Specification

-n, --name · string

Name of environment to export. If not specified, exports the currently active environment.

-p, --prefix · string

Full path to environment location (i.e. prefix) to export.

Output Options

-f, --file · string

File name or path for the exported environment.

This will silently overwrite any existing file of the same name in the current directory.

--format · string

Format for the exported environment. Available formats depend on installed plugins.

Common formats:

• yaml or environment-yaml - Standard YAML format (default)
• json or environment-json - JSON format

If not specified, format is determined by:

1. File extension (if --file is used)
2. Default to YAML

--json · boolean

For backwards compatibility: when used without --format and without --file, exports in JSON format.

If used with --file, only affects status messages (not the export format).

Channel Options

-c, --channel · string

Additional channel to include in the export. Can be specified multiple times.

-O, --override-channels · boolean

Do not include .condarc channels in the export.

--ignore-channels · boolean · default: false

Do not include channel names with package names in the export.

Use this when you want package specifications without channel prefixes (e.g., numpy instead of conda-forge::numpy).

Platform Options

--platform, --subdir · string

Target platform(s)/subdir(s) for export (e.g., linux-64, osx-64, win-64). Can be specified multiple times for multi-platform exports.

Multi-platform export is only supported by certain exporters (like environment-yaml).

--override-platforms · boolean

Override the platforms specified in the condarc.

Package Specification Options

--from-history · boolean · default: false

Build environment spec from explicit specs in history (packages you explicitly installed) rather than all packages.

Recommended for creating portable, minimal environment specifications. This exports only the packages you explicitly requested, not their dependencies.

--no-builds · boolean · default: false

Remove build specification from dependencies. This makes the export more portable across platforms.

Example:

• With builds: numpy=1.24.0=py311h1234567_0
• Without builds: numpy=1.24.0

Output Control

-v, --verbose · boolean

Can be used multiple times. Once for INFO, twice for DEBUG, three times for TRACE.

-q, --quiet · boolean

Do not display progress bar.

Export Formats

YAML Format (Default)

The standard conda environment format:

name: myenv
channels:
  - conda-forge
  - defaults
dependencies:
  - python=3.11.0=h1234567_0
  - numpy=1.24.0=py311h1234567_0
  - pandas=1.5.3=py311h1234567_0
  - pip:
    - requests==2.28.2
prefix: /home/user/miniconda3/envs/myenv

JSON Format

Structured JSON output:

{
  "name": "myenv",
  "channels": [
    "conda-forge",
    "defaults"
  ],
  "dependencies": [
    "python=3.11.0=h1234567_0",
    "numpy=1.24.0=py311h1234567_0"
  ],
  "prefix": "/home/user/miniconda3/envs/myenv"
}

Examples

Export current environment to stdout

conda env export

Exports the active environment in YAML format to the terminal.

Export to a file

conda env export --file environment.yml

Exports to environment.yml in the current directory.

Export specific environment

conda env export -n myenv --file myenv.yml

Exports the myenv environment to a file.

Export with minimal specification

conda env export --from-history --file environment.yml

Exports only explicitly installed packages (recommended for portability).

Export without build strings

conda env export --no-builds --file environment.yml

Exports without platform-specific build identifiers.

Export in JSON format

conda env export --format json --file environment.json

or for backwards compatibility:

conda env export --json > environment.json

Export without channels

conda env export --ignore-channels --file environment.yml

Exports packages without channel prefixes.

Export for multiple platforms

conda env export --platform linux-64 --platform osx-64 --file environment.yml

Creates a multi-platform environment file.

Export from history without builds

conda env export --from-history --no-builds --file environment.yml

Creates the most portable environment specification.

Common Use Cases

Share environment with team

Export a minimal, portable specification:

conda env export --from-history --no-builds --file environment.yml

Team members can recreate with:

conda env create -f environment.yml

Exact environment reproduction

Export with full specifications for exact reproduction:

conda env export --file environment-exact.yml

This includes all dependencies with exact versions and builds.

Cross-platform development

Export without build strings for cross-platform compatibility:

conda env export --no-builds --from-history --file environment.yml

Version control

Create minimal environment file for version control:

conda env export --from-history --no-builds --file environment.yml
git add environment.yml
git commit -m "Add environment specification"

Backup before major changes

Backup environment before updates:

conda env export --file backup-$(date +%Y%m%d).yml
conda update --all

CI/CD pipelines

Export for reproducible testing:

conda env export --from-history --no-builds --file ci-environment.yml

Export Strategies

Minimal Export

Best for: Sharing and version control

conda env export --from-history --no-builds

• Only explicitly installed packages
• No build strings
• Most portable

Complete Export

Best for: Exact reproduction

conda env export

• All packages including dependencies
• Full version and build specs
• Platform-specific

Cross-Platform Export

Best for: Multi-platform projects

conda env export --no-builds --from-history

• Works across different operating systems
• No platform-specific builds

Channel-Free Export

Best for: Flexibility

conda env export --ignore-channels --from-history

• No channel specifications
• Allows conda to choose best sources

Understanding Export Options

—from-history vs Full Export

Full export (default):

dependencies:
  - python=3.11.0=h1234567_0
  - numpy=1.24.0=py311h1234567_0
  - pandas=1.5.3=py311h1234567_0
  - pytz=2022.7.1=py311h1234567_0  # pandas dependency
  - python-dateutil=2.8.2=pyhd8ed1ab_0  # pandas dependency

With —from-history:

dependencies:
  - python=3.11
  - numpy
  - pandas

—no-builds Effect

With builds:

- numpy=1.24.0=py311h1234567_0

Without builds:

- numpy=1.24.0

Troubleshooting

Export contains too many packages

Use --from-history to export only explicitly installed packages:

conda env export --from-history --file environment.yml

Export not working across platforms

Remove build strings and use history:

conda env export --from-history --no-builds --file environment.yml

Unknown platform error

Ensure you’re specifying valid platforms:

conda env export --platform linux-64 --platform osx-64 --platform win-64

Valid platforms: linux-64, linux-aarch64, osx-64, osx-arm64, win-64

File already exists

The command will silently overwrite. To be safe:

# Backup first
cp environment.yml environment-backup.yml
# Then export
conda env export --file environment.yml

Best Practices

1. Version Control: Use --from-history --no-builds for files in version control
2. Documentation: Add comments to exported YAML files to explain special packages
3. Regular Updates: Export environment specifications regularly as your project evolves
4. Multiple Formats: Keep both minimal and complete exports for different purposes
5. Platform Specificity: Use --platform when you know the target platforms

Related Commands

• conda env create - Create environment from exported file
• conda env update - Update environment from specification
• conda list - List packages in environment
• conda env list - List all environments
• conda activate - Activate environment to export