QSBench Logo

QSBench

Ideal. Noisy. Reproducible.

High-quality synthetic quantum datasets with paired ideal and noisy expectation values. The definitive ground truth for Quantum Machine Learning research.

Featured Tool Demo Sandbox

Noise Robustness Benchmark

Predict noisy expectation values directly from circuit topology. Powered by the 4-part QSBench Demo suite.

Launch Benchmark
Preview

The Demo Suite

Four free-to-use datasets to validate your models before scaling up.

Core Demo

Balanced circuit families for general testing.

v1.0.0-demo

Depolarizing

Standard incoherent noise samples.

v1.0.0-demo

Amplitude

Realistic T1 relaxation noise simulation.

v1.0.0-demo

Transpilation

Hardware-aware circuit topology data.

v1.0.0-demo

Interactive QML Spaces

Explore functional demos trained on our free datasets.

Multi-Target Regression

Hugging Face Space Demo

Launch Demo →

Circuit Classifier

Hugging Face Space Demo

Launch Demo →

Entanglement Score

Hugging Face Space Demo

Launch Demo →

CNOT Estimator

Hugging Face Space Demo

Launch Demo →

Noise Detection

Hugging Face Space Demo

Launch Demo →

Complexity Clustering

Hugging Face Space Demo

Launch Demo →

Ready for Production?

Upgrade to our professional-grade datasets (up to 200,000 circuits, 10 qubits) for high-precision model training and publication-quality research.

Enter Full Catalog