TensorFlow Quantum

TensorFlow Quantum is Google’s open-source Python framework for hybrid quantum-classical machine learning, with 2,100+ stars on GitHub — it lets you build models that learn from quantum data using TensorFlow and Cirq together.

Overview

Quantum machine learning sits awkwardly between two toolchains: quantum circuits live in one library, gradient-based training in another, and gluing them together by hand is painful. TensorFlow Quantum (TFQ), actively maintained in 2026, is the bridge. It interleaves quantum algorithms and logic written in Cirq — Google’s quantum programming framework — with TensorFlow’s powerful, performant ML tooling, so a quantum circuit becomes a differentiable layer inside a Keras-style model. For execution it integrates with qsim for fast circuit simulation, letting you prototype hybrid models on classical hardware before targeting real quantum processors. TFQ focuses on modeling quantum data and near-term (NISQ-era) algorithms: variational circuits, quantum neural networks and hybrid classifiers are the natural workloads. It supports Python 3.10+, installs from PyPI, and comes with tutorials and documentation on tensorflow.org. The project is developed openly inside the tensorflow GitHub organization, alongside the main TensorFlow codebase it is designed to complement.

Key Features

• Hybrid models – quantum circuits as differentiable layers inside TensorFlow models.
• Cirq integration – define circuits with Google’s quantum programming framework.
• Fast simulation – qsim integration for high-performance circuit simulation.
• Quantum data focus – tooling designed for learning from quantum-generated data.
• TensorFlow ecosystem – Keras-style training, optimizers and deployment tooling included.
• Open and documented – Apache 2.0, on PyPI, with tutorials at tensorflow.org/quantum.

💳 Pricing

TensorFlow Quantum is free and open source under the Apache License 2.0. Install it from PyPI and run it on your own hardware; there is no paid edition, with costs limited to whatever compute you choose to use.

Who Is It For

Quantum computing researchers, ML engineers and advanced students exploring quantum machine learning who want quantum circuits and classical training loops in one framework.