Released version


Installing for Windows can be a little more involved: see the section below.

Brian is available on the Python package index:

It can therefore be installed using easy_install or pip (in newer versions, this needs the --pre flag to allow for the installation of a prelease):

easy_install brian2
pip install --pre brian2  # newer versions of pip
pip install brian2  # older versions of pip


In principle, the above commands also install Brian’s dependencies. Unfortunately, however, there are two circumstances where this doesn’t work:

  1. If numpy isn’t installed yet, it has to be installed in a separate step before any other dependencies (easy_install numpy or pip install numpy)
  2. On Python 2.x, installing sympy via easy_install does not work because it tries to install the Python 3 version (see this issue), use pip instead.

Alternatively, you can download the source package directly and uncompress it. You can then either run python install to install it, or simply add the source directory to your PYTHONPATH. Note that if you are using Python 3, directly running from the source directory is not possible, you have to install the package first.


We recommend first installing the Anaconda distribution of Python 2.7.

In order to get support for compiled C++ code (highly recommended for speed), you need to next install Visual Studio 2008 Express, which can be freely downloaded from Microsoft here. If you installed the 64 bit version of Python, then you now need to modify the file C:\Program Files (x86)\Microsoft Visual Studio 9.0\VC\vcvarsall.bat according to this patch to correct a bug in the free versions of Visual Studio.

Annoyingly, if you wish to use standalone mode on Windows, you will also need a more recent version of Visual Studio: any version of Visual Studio express from 2010 onwards (also freely available from Microsoft). You will need to apply a similar patch to the above.

For further reading on these issues which might help to get compiled C++ code running with a different Python installation (particularly for 64 bit Windows), see the following:

Development version

To run the latest development code, clone the git repository at github:

You can also use pip to install directly from github, use:

pip install

C extensions

During installation, Brian will try to compile a C++ version of the SpikeQueue, which increases the speed of synaptic propagation. If compilation fails, the pure Python version is used instead. Note that if you use the source package directly without an install, you have to trigger this compilation explicitly using python build_ext --inplace. When you are using the sources from github, this process additionally needs a working installation of Cython.

Testing Brian

If you have the nose testing utility installed, you can run Brian’s test suite:

import brian2

It should end with “OK”, possibly showing a number of skipped tests but no warnings or errors. For more control about the tests that are run see the developer documentation on testing.