vqubit

A dependency-free quantum computing framework for Python, written in V.

vqubit lets you build quantum circuits and run them on local simulators with zero third-party dependencies. The numerically heavy work — linear algebra, gate matrices and simulation — is written in V and compiled into a native extension; the public API is dependency-free Python on top of it.

The API mirrors Cirq closely, so if you know Cirq you already know vqubit.

import vqubit as vq

q0, q1 = vq.LineQubit.range(2)
circuit = vq.Circuit(vq.H(q0), vq.CNOT(q0, q1), vq.measure(q0, q1, key="result"))
print(circuit)

result = vq.Simulator().run(circuit, repetitions=100)
print(result.histogram(key="result"))

Highlights

• No dependencies. pip install vqubit installs nothing else. Everything runs offline, on your machine.
• Native speed. Linear algebra, gate matrices and simulation are compiled from V, not interpreted.
• One wheel, every Python. A single abi3 wheel covers CPython 3.11–3.15 per platform.
• Cross-platform. Linux and macOS (x86-64 and ARM).
• Cirq-compatible API. Build, transform and simulate circuits the way Cirq documents them.

Installation

$ pip install vqubit

Requires Python 3.11 or newer. No other packages are installed.

Documentation

The full documentation is published at https://rroblf01.github.io/vqubit/ in both Spanish and English. It starts from "what is a qubit?" and builds up to circuits and simulation.

Status

vqubit is alpha (0.1.0). It exports every one of cirq-core's top-level names, and the simulators, optimizers, transformers and devices are in place; see CHANGELOG.md for the full list. The API may still change before 1.0.

Windows support is currently blocked by the build tool vcraft, which does not yet produce Windows wheels.

Development

vqubit is built with vcraft. You need the V compiler, a C compiler, and uv for the development tooling.

$ python -m venv .venv && source .venv/bin/activate
$ pip install vcraft
$ uv pip install --group dev   # pytest, ruff, ty, build
$ vcraft develop               # build and install into the active virtualenv
$ python main.py               # run the example
$ pytest                       # run the test suite
$ vcraft build --release       # write an abi3 wheel into dist/

The dev and docs tooling are [dependency-groups] in pyproject.toml; use uv pip install --group docs for the documentation toolchain. (uv pip install adds to the virtualenv without pruning it — uv sync would remove the editable extension that vcraft develop installs.)

The development setup is the only thing that needs the V toolchain — the published wheels need nothing.

License

MIT. See LICENSE. © Ricardo Robles Fernández.