Installation

Requirements

• Python 3.10, 3.11, 3.12, or 3.13

• NumPy >= 1.20 (installed automatically)

MicroQuantum is pure Python + NumPy and is platform-independent at the package level. It is designed for Windows, Linux, macOS and BSD systems with a compatible Python + NumPy environment; see Platform Support for the CI-verified and manual verification status of each platform.

From PyPI

pip install microquantum

With uv:

uv pip install microquantum

Optional GPU acceleration (NumPy remains the default; CuPy is opt-in, and the gpu extra is only available on platforms where CuPy ships wheels):

pip install "microquantum[gpu]"

Verify the installation (on every platform):

python -c "import microquantum; print(microquantum.__version__)"

You should see 1.1.0. When the microquantum entry point is on your PATH:

microquantum --version

Linux

(CI-verified on GitHub Actions ``ubuntu-latest``)

Install a supported CPython on your distribution, typically already present or available from the package manager, for example:

# Debian / Ubuntu
sudo apt install python3 python3-venv python3-pip

# Fedora / RHEL / CentOS
sudo dnf install python3 python3-pip

Create and activate a virtual environment, then install:

python3 -m venv .venv
source .venv/bin/activate
python -m pip install microquantum

Or install into the user site without a virtual environment:

python3 -m pip install --user microquantum

Windows

(CI-verified on GitHub Actions ``windows-latest``)

Install a supported CPython for Windows from python.org (or the Windows Store), then use the Microsoft Store / python.org python launcher py to create a virtual environment:

py -m venv .venv
.venv\Scripts\Activate.ps1
python -m pip install microquantum

In cmd.exe the activation script is .venv\Scripts\activate.bat instead. Installing when the environment is active places the microquantum.exe console script on PATH, so verify with:

python -c "import microquantum; print(microquantum.__version__)"
microquantum --version

macOS

(CI-verified on GitHub Actions ``macos-latest``)

Install a supported CPython for macOS, for example with Homebrew:

brew install python

Then create and activate a virtual environment and install:

python3 -m venv .venv
source .venv/bin/activate
python -m pip install microquantum

Apple Silicon and Intel macOS are covered by the CI runners used; the SDK is pure Python + NumPy so it uses the prebuilt NumPy wheels for the installed architecture.

BSD

(Not CI-verified; supported by design — see :doc:`/releases/platform-support`)

On a BSD system (such as FreeBSD or OpenBSD) install a supported CPython from the ports/packages system, for example:

# FreeBSD
sudo pkg install python311 py311-pip

# OpenBSD
doas pkg_add python-3.11 py3-pip

Then create a virtual environment and install with pip:

python3.11 -m venv .venv
source .venv/bin/activate
python -m pip install microquantum

MicroQuantum is portable Python + NumPy and loads pure Python wheels (py3-none-any), so no BSD-specific binaries are needed. NumPy must be able to install on the BSD release in use; a source build is used where no wheel is available for the platform/architecture.

From source

git clone https://github.com/ajit-ai/microquantum.git
cd microquantum
pip install -e .

Development tooling (tests, lint, type check, docs) is managed with uv:

uv sync --group dev

Documentation build extras:

pip install "microquantum[docs]"

Next: Quickstart.