Contact us:

Conda Package Installation (Linux, Mac OS X 10.10 or later, Win64)

This method relies on the Continuum Analytics Anaconda environment, which must be installed first.

NOTE: If you have a system-wide Anaconda installation (most people do not), then you should be aware of a bug in conda:

conda install -c sys-bio roadrunner

Debian Package Installation (Tested on Ubuntu 14.04 and Debian 8)

Add the following line to /etc/apt/sources.list:

deb ./

Then run these commands:

sudo apt-get update
sudo apt-get install libroadrunner-python # will warn about authentication, it's okay

Installation from Source

This page only describes the installation of libRoadRunner. To install the fully integrated environment that includes libRoadRunner, Antimony, Matlab exporter etc, please follow the Tellurium link.

Jump to:

Installing  on Windows
Installing on the Mac or Linux


There is one distribution for Windows that includes the Python bindings and the DLLs you can use to include libRoadRunner in your own applications. For Python use we recommend the Tellurium distribution as this includes a fully automated installer and doesn’t interfere with your existing Python setup.

  1. Download & Run:
    • Download the Windows binaries
    • Unzip and copy the contents of the site-packages directory to your Python site-packages.
  2. Simulate a model in your python console (This invokes the  provided model feedback.xml):
    import roadrunner.testing
    r = roadrunner.testing.getRoadRunner("feedback.xml")

    For one of your own models you could use:

    import roadrunner
    r = roadrunner.RoadRunner ("mymodel.xml")
  3. Follow Basic Tutorial in Python API documentation.

If you want more control over the installation we provide advanced instructions 

LibRoadRunner Binaries

The bin directory in the zip file contains the Windows dlls that you can call from other applications. We provide a C++ API and C API. Almost all programming languages have some way to call C functions, this allows the libroadrunner C API to be used to many other situations, for example, calling libRoadRunner form Object Pascal, C#, R, Julia, …. We refer you to the C API documentation for details.

The following binary is the 32-bit version.

Download the binaries here (Look for

Mac OS X and Linux

We provide complete, self-contained binary python packages for Mac OSX and Linux which are designed to work with the built in system Python.

The cpplibroadrunner-1.2.0-xxx files contain the libRoadRunner header files, shared libraries and command line programs and sbml test suite. To use these, just extract these, and the ‘libroadrunner/bin’ directory to your path and the ‘libroadrunner/lib’ directory your library path (LD_LIBRARY_PATH on Linux, DYLD_LIBRARY_PATH on OSX).

System requirements are:

Mac OSX:

Any 64 bit system running Mac OSX 10.6 and above. LibRoadRunner is designed to work with the default system Python on Mac OSX 10.6 and above. It will also work if users have upgraded to a newer Python and / or Numpy.


Any 32 or 64 bit intel system with glibc >= 2.1.2, Python >= 2.6 and Numpy >= 1.2.1. Basically this means that any Linux system newer than 2006 should work fine, provided that the minimum Python libraries are installed. It should work out of the box on Ubuntu >= 9.04, Fedora >= 11, Suse >= 11 or RHEL 5 with upgraded Python.


Download and unzip the
OS X binaries Bring up the Mac terminal, cd to where you unzipped the files. Then from the terminal run either:

1: install for only current user

python install --user

2: or install system wide (note requires root privilege)

sudo python install

Though not required, it is recommended to ipython and matplotlib for interaction simulation and plotting. After installing, one may verify LibRoadRunner by running the following commands in python:

>>>import roadrunner
>>> roadrunner.__version__ # displays the version
'1.1.0; Compiler: clang "3.3 (tags/RELEASE_33/final)"; Date: Mar 15 2014, 13:15:29'
# import the built in unit tests and run them 
>>> import roadrunner.testing 
>>> roadrunner.testing.runTester()

You should see a series of lines displaying what test is run, and the result which should be either PASS or NO TEST. A graphical install is in development for Mac OSX.

Follow Basic Tutorial in Python API documentation.