About

The Quantum Benchmark Zoo is a catalog of protocols for measuring quantum computer performance. Quantum benchmarking is fragmented across papers, vendor documentation, and software repositories; this site collects the protocols in one place, with a consistent structure: what each benchmark measures, how it works, its strengths and limitations, key papers, and reference implementations.

How entries are organized

Every benchmark is filed under one of seven categories:

• Component-level: Characterize individual gates, qubits, and operations in isolation: error rates, coherence, and calibration quality. Home of the randomized benchmarking family.
• System-level: Exercise a whole processor with structured or random circuits to produce holistic scores that reflect qubit count, fidelity, connectivity, and the compiler together.
• Application-level: Measure end-to-end performance on programs representative of real workloads, from algorithm subroutines to full application suites.
• Error correction: Score the error-corrected layer: how fast logical errors fall as codes scale, what logical qubits and fault-tolerant primitives cost, and whether decoders keep up.
• Software stack: Benchmark the classical software around the QPU (compilers, transpilers, SDKs, and verification tools) on circuit corpora with known baselines or optima.
• Platform-specific: Benchmarks built for hardware outside the digital gate model: quantum annealers, photonic samplers, and analog quantum simulators.
• Characterization: Diagnostic protocols (tomography, fidelity estimation, noise learning) that reconstruct what a device actually does: the toolbox that score-style benchmarks build on.

Entries aim for a neutral, cite-your-sources tone: claims about what a benchmark shows, and where it falls short, should trace back to the literature.

Scope

An entry is a named benchmark, protocol, or suite with a defined procedure that yields a score or diagnostic: the kind of thing a paper or spec introduces and others can reproduce. A few adjacent kinds of things are deliberately not catalogued:

• Bare device metrics: T1, T2, single-number gate or readout fidelities, and qubit counts. They are measurement results, not protocols; the protocols that produce them (randomized benchmarking, tomography) are catalogued instead.
• Results platforms and leaderboards: e.g. Metriq, which tracks benchmark results over time rather than defining benchmarks.
• Datasets and trackers without a scoring protocol, with the exception of problem libraries (such as HamLib) that the field uses directly as benchmark sets.
• Benchmarks of AI coding assistants on quantum programming tasks (Qiskit HumanEval and friends): they measure language models, not quantum computers or their software stacks.

Further reading

Four recent surveys anchor the field and informed this catalog's coverage:

• Benchmarking quantum computers (Proctor, Young, Baczewski & Blume-Kohout, Nature Reviews Physics 7, 105–118, 2025).
• A Practical Introduction to Benchmarking and Characterization of Quantum Computers (Hashim et al., PRX Quantum 6, 030202, 2025).
• Quantum computer benchmarking: a collection of metrics (Lall et al., the NPL-led QCMet collection, 2025).
• A systematic literature review of quantum benchmarks (Rohe et al., 2025), covering 329 studies.

Contributing

Each benchmark is a single Markdown file with structured frontmatter, validated against a schema at build time, so a malformed entry fails the build rather than shipping. To add a benchmark: copy an existing file in src/content/benchmarks/, fill in the frontmatter and body, and open a pull request on GitHub. The repository's CONTRIBUTING.md has the full field reference.

Related resources

• Metriq: a Unitary Foundation platform for submitting and tracking quantum benchmark results over time.
• Error Correction Zoo: a catalog of classical and quantum error-correcting codes.
• Quantum Algorithm Zoo: a catalog of quantum algorithms and their speedups.