Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/prefix-dev/pixi/llms.txt

Use this file to discover all available pages before exploring further.

The pixi import command imports dependencies from external environment files (conda environment.yml or requirements.txt) into your pixi workspace.

Usage

pixi import [OPTIONS] <FILE>

Arguments

FILE
string
required
Path to the environment file to import. Supported formats:
  • environment.yml or environment.yaml (conda environment file)
  • requirements.txt (PyPI requirements file)
If --format is not specified, pixi will try to detect the format automatically.

Options

Format Selection

--format
string
Explicitly specify the file format to interpret. Options:
  • conda-env - Conda environment file (environment.yml)
  • pypi-txt - PyPI requirements file (requirements.txt)
Case-insensitive.
pixi import environment.yml --format conda-env
pixi import requirements.txt --format pypi-txt

Platform Configuration

--platform
Platform
Specify platforms for the imported environment. Can be specified multiple times.Short flag: -p
pixi import environment.yml --platform linux-64 --platform osx-arm64

Environment and Feature Names

--environment
string
Name for the created environment.Short flag: -eIf not provided, attempts to use the name field from conda environment files. For PyPI files, this is required.
--feature
string
Name for the created feature.Short flag: -fIf not provided, uses the same value as --environment.
pixi import environment.yml --environment ml-env --feature ml
pixi import requirements.txt --environment test --feature testing

Import Formats

Conda Environment Files

When importing conda environment files, pixi extracts:
  • Dependencies (both conda and pip)
  • Channels
  • Environment name (if specified)
Example environment.yml:
name: myenv
channels:
  - conda-forge
  - defaults
dependencies:
  - python=3.11
  - numpy>=1.20
  - pip:
    - requests
    - flask
Import command:
pixi import environment.yml
Resulting pixi.toml:
[feature.myenv.dependencies]
python = "3.11.*"
numpy = ">=1.20"

[feature.myenv.pypi-dependencies]
requests = "*"
flask = "*"

[environments]
myenv = ["myenv"]

PyPI Requirements Files

Unnamed requirements (direct URLs without package names) are currently unsupported. Constraints in requirements files are ignored with a warning.
Example requirements.txt:
requests==2.28.0
flask>=2.0,<3.0
django
Import command:
pixi import requirements.txt --environment web
Resulting pixi.toml:
[feature.web.pypi-dependencies]
requests = "==2.28.0"
flask = ">=2.0,<3.0"
django = "*"

[environments]
web = ["web"]

Behavior

Automatic Format Detection

If --format is not specified, pixi tries each supported format in order:
  1. Conda environment format
  2. PyPI requirements format
If all formats fail, an error is returned:
pixi import unknown-file.txt
# Error: Tried all formats for input file, but none were successful.
#        Pass a `format` argument to see the specific error for that format.

Environment Creation

  • New environment: Creates a new environment with the specified name and feature
  • Existing environment: Adds the feature to the existing environment if not already present
  • Feature addition: If the environment exists but doesn’t include the feature, the feature is added to it

Platform Handling

If platforms are specified with --platform, they are added to the feature:
pixi import environment.yml --platform linux-64 --platform osx-64
This adds the platforms to the feature’s platform list.

Examples

Import Conda Environment

pixi import environment.yml
Uses the name field from environment.yml:
name: data-science
channels:
  - conda-forge
dependencies:
  - python=3.11
  - pandas
  - numpy

Import with Custom Names

pixi import environment.yml --environment ml --feature machine-learning
Creates an environment named “ml” with a feature named “machine-learning”.

Import PyPI Requirements

pixi import requirements.txt --environment backend --feature api
PyPI files require explicit --environment or --feature specification.

Import with Platforms

pixi import environment.yml --platform linux-64 --platform osx-arm64
Imports dependencies and configures them for specific platforms.

Import with Explicit Format

pixi import deps.txt --format pypi-txt --environment deps
Forces interpretation as a PyPI requirements file.

Import to Existing Environment

# First import
pixi import base.yml --environment prod

# Add more dependencies to the same environment
pixi import additional.yml --environment prod --feature extras
The feature “extras” is added to the existing “prod” environment.

Global Options

--manifest-path
string
Path to pixi.toml, pyproject.toml, or workspace directory.Short flag: -m
pixi import environment.yml --manifest-path /path/to/project

Config Options

These options override global configuration:
  • --tls-no-verify: Disable TLS verification
  • --auth-file <FILE>: Path to authentication file
  • --pypi-keyring-provider <PROVIDER>: Keyring provider for PyPI

Troubleshooting

Missing Environment Name

Error: Missing name: provide --feature or --environment, or set 'name:' in input file Solution: Specify the environment/feature name explicitly:
pixi import environment.yml --environment myenv
Or ensure the conda environment file has a name field.

Unsupported Format

Error: Tried all formats for input file, but none were successful Solution: Specify the format explicitly:
pixi import file.txt --format pypi-txt

Unnamed Requirements

Error: Error parsing input file: unnamed requirements are currently unsupported Solution: Ensure all requirements in requirements.txt have package names:
# ❌ Unsupported
https://github.com/user/repo/archive/main.zip

# ✅ Supported
mypackage @ https://github.com/user/repo/archive/main.zip

Constraints Ignored

Warning: Constraints detected in input file, but these are currently unsupported Constraints (lines starting with -c in requirements.txt) are not yet supported and will be ignored.

Limitations

Current limitations:
  • Unnamed requirements are not supported
  • Constraints files (-c constraints.txt) are ignored
  • Environment variables from conda files are not imported

See Also

Build docs developers (and LLMs) love