Use a Python.org universal install binary. Install as many versions as you want. You will always be able to get a particular version by typing, e.g., `python3.1

Use a Python.org universal install binary. Install as many versions as you want. You will always be able to get a particular version by typing, e.g., `python3.10`. Then use that version to create your virtual environment: `python3.10 -m venv env`. You can start your virtual environment with `source env/bin/activate`. Here is 3.8: https://www.python.org/downloads/release/python-3810/ https://www.python.org/downloads/release/python-3810/ Here is 3.9: https://www.python.org/downloads/release/python-3913/ https://www.python.org/downloads/release/python-3913/ Here is 3.10: https://www.python.org/downloads/release/python-3108/ https://www.python.org/downloads/release/python-3108/ Here is 3.11: https://www.python.org/downloads/release/python-3114/ https://www.python.org/downloads/release/python-3114/ Whatever you do, do not use Homebrew, which has a habit of breaking your python environment when it updates patch versions. https://justinmayer.com/posts/homebrew-python-is-not-for-you/ https://justinmayer.com/posts/homebrew-python-is-not-for-you.... Pyenv is usually ok, but sometimes you need to be a shell scripting wizard. And I am not that. Conda is probably ok too. If you are in a data science ecosystem, go with Conda and you will be fine. And building your own has always been fine.