Conda package specification#

A conda package is an archive file that contains:

  • Metadata under the info/ directory.

  • A collection of files that are installed directly into an install prefix.

There are currently two formats of archives that are supported:




The original format of conda packages. Is the default output of conda-build.


2nd Gen. This is a more compact and thus faster. Can be outputed from conda-build by setting output in .condarc file.

The formats are identical across platforms and operating systems. During the install process, all files are extracted into the install prefix, with the exception of the ones in info/. Installing a conda package into an environment is similar to executing the following commands:

cd <environment prefix>
tar xjf some-package-1.0-0.tar.bz2

Only files, including symbolic links, are part of a conda package. Directories are not included. Directories are created and removed as needed, but you cannot create an empty directory from the tar archive directly.

Package metadata#

The info/ directory contains all metadata about a package. Files in this location are not installed under the install prefix. Although you are free to add any file to this directory, conda only inspects the content of the files discussed below.


This file contains basic information about the package, such as name, version, build string, and dependencies. The content of this file is stored in repodata.json, which is the repository index file, hence the name index.json. The JSON object is a dictionary containing the keys shown below. The filename of the conda package is composed of the first 3 values, as in: <name>-<version>-<build>.tar.bz2 or <name>-<version>-<build>.conda.






The lowercase name of the package. May contain the "-" character.



The package version. May not contain "-". Conda acknowledges PEP 440.



The build string. May not contain "-". Differentiates builds of packages with otherwise identical names and versions, such as:

  • A build with other dependencies, such as Python 3.4 instead of Python 2.7.

  • A bug fix in the build process.

  • Some different optional dependencies, such as MKL versus ATLAS linkage. Nothing in conda actually inspects the build string. Strings such as np18py34_1 are designed only for human readability and conda never parses them.



A non-negative integer representing the build number of the package.

Unlike the build string, the build_number is inspected by conda.

Conda uses it to sort packages that have otherwise identical names and versions to determine the latest one.

This is important because new builds that contain bug fixes for the way a package is built may be added to a repository.


list of strings

A list of dependency specifications, where each element is a string, as outlined in Package match specifications.



Optional. The architecture the package is built for.

EXAMPLE: x86_64

Conda currently does not use this key.



Optional. The OS that the package is built for.


Conda currently does not use this key.

Packages for a specific architecture and platform are usually distinguished by the repository subdirectory that contains them.

See Repository structure and index.


Lists all files that are part of the package itself, 1 per line. All of these files need to get linked into the environment. Any files in the package that are not listed in this file are not linked when the package is installed. The directory delimiter for the files in info/files should always be "/", even on Windows. This matches the directory delimiter used in the tarball.


Optional file. Lists all files that contain a hard-coded build prefix or placeholder prefix, which needs to be replaced by the install prefix at installation time.


Due to the way the binary replacement works, the placeholder prefix must be longer than the install prefix.

Each line of this file should be either a path, in which case it is considered a text file with the default placeholder /opt/anaconda1anaconda2anaconda3, or a space-separated list of placeholder, mode, and path, where:

  • Placeholder is the build or placeholder prefix.

  • Mode is either text or binary.

  • Path is the relative path of the file to be updated.

EXAMPLE: On Windows:

"C:\Users\username\anaconda\envs\_build" text "Scripts/script2.bat"
"C:/Users/username/anaconda/envs/_build" binary "Scripts/binary"

EXAMPLE: On macOS or Linux:

/Users/username/anaconda/envs/_build binary bin/binary
/Users/username/anaconda/envs/_build text share/text


The directory delimiter for the relative path must always be "/", even on Windows. The placeholder may contain either "\" or "/" on Windows, but the replacement prefix will match the delimiter used in the placeholder. The default placeholder /opt/anaconda1anaconda2anaconda3 is an exception, being replaced with the install prefix using the native path delimiter. On Windows, the placeholder and path always appear in quotes to support paths with spaces.


Optional file. The software license for the package.


Optional file. Contains the entries in the About section of the meta.yaml file. The following keys are added to info/about.json if present in the build recipe:

  • home.

  • dev_url.

  • doc_url.

  • license_url.

  • license.

  • summary.

  • description.

  • license_family.


A directory containing the full contents of the build recipe.


The fully rendered build recipe. See conda render.

This directory is present only when the the include_recipe flag is True in the Build section.

Repository structure and index#

A conda repository---or channel---is a directory tree, usually served over HTTPS, which has platform subdirectories, each of which contains conda packages and a repository index. The index file repodata.json lists all conda packages in the platform subdirectory. Use conda index to create such an index from the conda packages within a directory. It is simple mapping of the full conda package filename to the dictionary object in info/index.json described in Adding pre-link, post-link, and pre-unlink scripts.

In the following example, a repository provides the conda package misc-1.0-np17py27_0.tar.bz2 on 64-bit Linux and 32-bit Windows:

<some path>/linux-64/repodata.json


Both conda packages have identical filenames and are distinguished only by the repository subdirectory that contains them.

Package match specifications#

This match specification is not the same as the syntax used at the command line with conda install, such as conda install python=3.4. Internally, conda translates the command line syntax to the spec defined in this section.

EXAMPLE: python=3.4 is translated to python 3.4.*. conda search 'python=3.1' does NOT bring up Python 3.10, only Python 3.1.*.

Package dependencies are specified using a match specification. A match specification is a space-separated string of 1, 2, or 3 parts:

  • The first part is always the exact name of the package.

  • The second part refers to the version and may contain special characters. See table below.

  • The third part is always the exact build string. When there are three parts, the second part must be the exact version.

Version Special Characters#




<, >, <=, >=

Relational operators on versions, which are compared using PEP-440.

<=1.0 matches 0.9, 0.9.1, and 1.0, but not 1.0.1.

==, and !=

Exact equality and not equalities.

==0.5.1 matches 0.5.1 and not anything else while !=0.5.1 matches everything but.


Compatibility Release

~=0.5.3 is equivalent to >=0.5.3, <0.6.0a



1.0|1.2 matches version 1.0 or 1.2.


Matches 0 or more characters in the version string. In terms of regular expressions, it is the same as r'.*'.

1.0|1.4* matches 1.0, 1.4 and 1.4.1b2, but not 1.2.



>=2,<3 matches all packages in the 2 series. 2.0, 2.1, and 2.9 all match, but 3.0 and 1.0 do not.


, has higher precedence than |, so >=1,<2|>3 means greater than or equal to 1 AND less than 2 or greater than 3, which matches 1, 1.3 and 3.0, but not 2.2.


For package match specifications, pre-release versioning is also supported such that >1.0b4 will match 1.0b5 and 1.0rc1 but not 1.0b4 or 1.0a5.

Conda parses the version by splitting it into parts separated by |. If the part begins with <, >, =, or !, it is parsed as a relational operator. Otherwise, it is parsed as a version, possibly containing the "*" operator.

Remember that the version specification cannot contain spaces, as spaces are used to delimit the package, version, and build string in the whole match specification. python >= 2.7 is an invalid match specification. However, "python >= 2.7" (with double or single quotes) is matched as any version of a package named python>=2.7.

Examples of Package Specs#

The build string constraint "numpy=1.11.2=*nomkl*" matches the NumPy 1.11.2 packages without MKL, but not the normal MKL NumPy 1.11.2 packages.

The build string constraint "numpy=1.11.1|1.11.3=py36_0" matches NumPy 1.11.1 or 1.11.3 built for Python 3.6, but not any versions of NumPy built for Python 3.5 or Python 2.7.

The following are all valid match specifications for numpy-1.8.1-py27_0:

  • numpy

  • numpy 1.8*

  • numpy 1.8.1

  • numpy >=1.8

  • numpy ==1.8.1

  • numpy 1.8|1.8*

  • numpy >=1.8,<2

  • numpy >=1.8,<2|1.9

  • numpy 1.8.1 py27_0

  • numpy=1.8.1=py27_0

Command Line Match Spec Examples#

When using the command line, put double or single quotes around any package version specification that contains the space character or any of the following characters: <, >, *, or |.




conda install numpy=1.11

The fuzzy constraint numpy=1.11 matches 1.11, 1.11.0, 1.11.1, 1.11.2, 1.11.18, and so on.

conda install numpy==1.11

The exact constraint numpy==1.11 matches 1.11, 1.11.0,, and so on.

conda install "numpy=1.11.1|1.11.3"

The OR constraint "numpy=1.11.1|1.11.3" matches with 1.11.1 or 1.11.3.

conda install "numpy>1.11"

Any numpy version 1.12.0a or greater.

conda install "numpy>=1.8,<2"

The AND constraint "numpy>=1.8,<2" matches with 1.8 and 1.9 but not 2.0.