Before installing the library, create a virtual environment for your project using:

$ python3 -m venv ~/luma-env

This creates a virtual environment in the home directory called luma-env and a Python executable at ~/luma-env/bin/python.

Next, install the latest version of the library in the virtual environment with:

$ ~/luma-env/bin/python -m pip install --upgrade luma.lcd

This will normally retrieve all of the dependencies luma.lcd requires and install them automatically.

Installing Dependencies

If pip is unable to automatically install its dependencies you will have to add them manually. To resolve the issues you will need to add the appropriate development packages to continue.

If you are using Raspbian or Raspberry Pi OS you should be able to use the following commands to add the required packages:

$ sudo apt-get update
$ sudo apt-get install python3 python3-pip python3-pil libjpeg-dev zlib1g-dev libfreetype6-dev liblcms2-dev libopenjp2-7 libtiff5 -y

If you are not using Raspberry Pi OS or Raspbian you will need to consult the documentation for your Linux distribution to determine the correct procedure to install the dependencies.


If your distribution includes a pre-packaged version of Pillow, use it instead of installing from pip. On many distributions the correct package to install is python3-imaging. Another common package name for Pillow is python3-pil.:

$ sudo apt-get install python3-imaging


$ sudo apt-get install python3-pil


luma.lcd uses hardware interfaces that require permission to access. After you have successfully installed luma.lcd you may want to add the user account that your luma.lcd program will run as to the groups that grant access to these interfaces.:

$ sudo usermod -a -G spi,gpio,i2c pi

Replace pi with the name of the account you will be using.