NiBabel is a pure Python package at the moment, and it should be easy to get NiBabel running on any system. For the most popular platforms and operating systems there should be packages in the respective native packaging format (DEB, RPM or installers). On other systems you can install NiBabel using pip or by downloading the source package and running the usual python install.

Installer and packages

pip and the Python package index

If you are not using a Linux package manager, then best way to install NiBabel is via pip. If you don’t have pip already, follow the pip install instructions.

Then open a terminal ( on OSX, cmd or Powershell on Windows), and type:

pip install nibabel

This will download and install NiBabel.

If you really like doing stuff manually, you can install NiBabel by downoading the source from NiBabel pypi . Go to the pypi page and select the source distribution you want. Download the distribution, unpack it, and then, from the unpacked directory, run:

pip install .

If you get permission errors, this may be because pip is trying to install to the system directories. You can solve this error by using sudo, but we strongly suggest you either do an install into your “user” directories, like this:

pip install --user .

or you work inside a virtualenv.


Our friends at NeuroDebian have packaged NiBabel at NiBabel NeuroDebian. Please follow the instructions on the NeuroDebian website on how to access their repositories. Once this is done, installing NiBabel is:

apt-get update
apt-get install python-nibabel

Install a development version

If you want to test the latest development version of nibabel, or you’d like to help by contributing bug-fixes or new features (excellent!), then this section is for you.


  • Python 2.7 or greater
  • NumPy 1.5 or greater
  • SciPy (optional, for full SPM-ANALYZE support)
  • PyDICOM 0.9.7 or greater (optional, for DICOM support)
  • Python Imaging Library (optional, for PNG conversion in DICOMFS)
  • nose 0.11 or greater (optional, to run the tests)
  • mock (optional, to run the tests)
  • sphinx (optional, to build the documentation)

Get the development sources

You can download a tarball of the latest development snapshot (i.e. the current state of the master branch of the NiBabel source code repository) from the NiBabel github page.

If you want to have access to the full NiBabel history and the latest development code, do a full clone (AKA checkout) of the NiBabel repository:

git clone


Just install the modules by invoking:

pip install .

See pip and the Python package index for advice on what to do for permission errors.

Validating your install

For a basic test of your installation, fire up Python and try importing the module to see if everything is fine. It should look something like this:

Python 2.7.8 (v2.7.8:ee879c0ffa11, Jun 29 2014, 21:07:35)
[GCC 4.2.1 (Apple Inc. build 5666) (dot 3)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
>>> import nibabel

To run the nibabel test suite, from the terminal run nosetests nibabel or python -c "import nibabel; nibabel.test().

To run an extended test suite that validates nibabel for long-running and resource-intensive cases, please see Advanced Testing.