In quantum programs, Bugs4Q is a widely used benchmark containing real quantum defects. However, its evaluation assumes that benchmark labels remain valid and that generated fixes execute in the target environment. We evaluate two Bugs4Q versions containing 67 unique real Qiskit defects, adding executable tests where missing, and re-run all entries across six pinned Qiskit releases (0.25.0, 0.45.0, 1.0.0, 1.1.1, 2.0.0, and 2.3.1). We find that quantum benchmarks can suffer from silent label inversion: entries become invalid without errors when reference fixes stop executing or buggy programs no longer reproduce failures. Thus, correctness depends on the (benchmark, version) pair rather than the benchmark alone. We evaluate four LLMs (GPT-4o-mini, GPT-5o-mini, GPT-5.4, and GPT-5.4-mini), generating up to 10 repair candidates per defect and testing them across all versions. GPT-5.4 achieves the highest pass@10 (48.8%), followed by GPT-5.4-mini (47.3%), GPT-5o-mini (30.3%), and GPT-4o-mini (22.6%). All models perform best on Qiskit 0.45.0 and decline after the Qiskit 1.0 transition. Many failures arise from deprecated or incompatible APIs rather than incorrect repairs, and 64\% of successful repairs occur on entries invalid under the target version. We release a re-validated, version-pinned Bugs4Q benchmark and show that benchmark validation must precede repair evaluation.
In quantum programs, Bugs4Q is a widely used benchmark containing real quantum defects. However, its evaluation assumes that benchmark labels remain valid and that generated fixes execute in the target environment. We evaluate two Bugs4Q versions containing 67 unique real Qiskit defects, adding executable tests where missing, and re-run all entries across six pinned Qiskit releases (0.25.0, 0.45.0, 1.0.0, 1.1.1, 2.0.0, and 2.3.1). We find that quantum benchmarks can suffer from silent label inversion: entries become invalid without errors when reference fixes stop executing or buggy programs no longer reproduce failures. Thus, correctness depends on the (benchmark, version) pair rather than the benchmark alone. We evaluate four LLMs (GPT-4o-mini, GPT-5o-mini, GPT-5.4, and GPT-5.4-mini), generating up to 10 repair candidates per defect and testing them across all versions. GPT-5.4 achieves the highest pass@10 (48.8%), followed by GPT-5.4-mini (47.3%), GPT-5o-mini (30.3%), and GPT-4o-mini (22.6%). All models perform best on Qiskit 0.45.0 and decline after the Qiskit 1.0 transition. Many failures arise from deprecated or incompatible APIs rather than incorrect repairs, and 64% of successful repairs occur on entries invalid under the target version. We release a re-validated, version-pinned Bugs4Q benchmark and show that benchmark validation must precede repair evaluation.
Large language models can generate plausible quantum code, but it is unclear whether they can reliably target the specific software development kit (SDK) version requested by the user. We study this problem as API drift and introduce quantum-api-drift, a benchmark for measuring version fidelity, defined here as execution success on the requested SDK version, cross-version compatibility, failure modes, and documentation-guided repair in LLM-generated quantum SDK code. We instantiate the benchmark with Qiskit, a representative quantum SDK that underwent substantial interface changes across v0.43, v1.3, and v2.0. We evaluate 17 models on 50 tasks with 3 samples per prompt, yielding 450 generated samples and 1,350 executions per model. Sixteen models are tested in a matched REST API setting with a 1024-token output cap, while GPT-5.4 (Codex CLI) is reported separately as a reference configuration. Across the 16 matched REST models, diagonal Pass@1 ranges from 0.02 to 0.85. Claude Opus 4.7 is strongest on v0.43 and v2.0, while Grok 4.20 is strongest on v1.3 at 0.513. Error profiles differ systematically by model strength: weaker models fail mainly with broken imports, while stronger models more often reach deprecation-level failures. Documentation-guided repair succeeds for 0.19 to 0.59 of repair attempts overall and is consistently much more effective for migration to v2.0 than to v1.3. The benchmark artifacts are publicly available at https://github.com/arasyi/quantum-api-drift. These results show that version alignment is a distinct evaluation axis for quantum code generation and that API drift remains only partly recoverable even with migration guidance.
Large language models (LLMs) can generate quantum programs, but reliable evaluation remains difficult because benchmarks differ in task composition, scoring, and failure handling. We evaluate 15 hosted frontier, open-weight, quantum-specialized, and retrieval-augmented models on five quantum code-generation suites (530 tasks), plus 10 models on our new QuantumBenchEval (QBE) benchmark (96 tasks across six topics). We report bootstrap confidence intervals, multiplicity-adjusted paired comparisons, and audits of output validity and evaluator behavior. QiskitHumanEval-Standard approaches saturation (98.7% pass@1), while its Hard variant retains headroom (63.8%). None of the five top-two comparisons across primary suites remains significant after multiple-comparison correction. Evaluator design substantially changes scores: removing checks for unspecified quantities raises GPT-6 Astra's QBE pass@1 from 0% to 94.1% (T1) and from 1.2% to 92.9% (T6), without changing model outputs. We identify 17 QuanBench-117 tasks with shared failures linked to benchmark-owned imports. Controlled probes also reveal unexecuted unit tests, leakage of canonical functions into candidate execution, and reference imports that block otherwise valid candidates. These defects can produce both false failures and false passes, so historical QuanBench results require revalidation under an isolated evaluator. Overall, leaderboard scores alone do not adequately characterize progress in quantum code generation. We release our harness, per-sample results, and audit artifacts and recommend validating evaluators, isolating candidate code, and reporting uncertainty and output validity.
Realistic benchmarks of reproducible bugs and fixes are vital to good experimental evaluation of debugging and testing approaches. However, there is no suitable benchmark suite that can systematically evaluate the debugging and testing methods of quantum programs until now. This paper proposes Bugs4Q, a benchmark of thirty-six real, manually validated Qiskit bugs from four popular Qiskit elements (Terra, Aer, Ignis, and Aqua), supplemented with the test cases for reproducing buggy behaviors. Bugs4Q also provides interfaces for accessing the buggy and fixed versions of the Qiskit programs and executing the corresponding test cases, facilitating the reproducible empirical studies and comparisons of Qiskit program debugging and testing tools. Bugs4Q is publicly available at https://github.com/Z-928/Bugs4Q
Large language models (LLMs) have shown strong results in classical code generation and are increasingly used to write the unit tests that safeguard that code. Whether they can do the same for quantum software is less clear, because quantum programs are probabilistic, depend on numerical tolerances, and are judged by domain-specific checkers, so a well-formed test can still disagree with the benchmark's verdict. Existing quantum benchmarks measure whether a model produces a correct solution, but not whether its tests can distinguish correct implementations from faulty ones. We introduce QTestBench, a framework that evaluates LLM-generated quantum test suites against independent benchmark checkers, never treating the tests themselves as ground truth. For each task, QTestBench prompts models to produce syntactic, semantic, and behavioral tests, injects candidate solutions at runtime, and compares each test verdict with the checker's label under three validation cohorts, each answering a distinct diagnostic question. We benchmark five LLMs on eight PennyLane QHack challenges and ten Qiskit HumanEval tasks. Generated suites almost always execute, but discrimination lags behind: the best generator reaches a test quality of 0.71 on QHack and 0.34 on Qiskit HumanEval, so executability alone overstates test quality. Oracle-verified synthetic mutants stabilize bug-detection estimates when natural faulty solutions are scarce, as on three of ten Qiskit tasks that yield none. QTestBench offers a reproducible, cross-framework benchmark for the trustworthiness of LLM-generated quantum tests.
The reproducibility of software defect datasets is essential for obtaining reliable and comparable research results. Zhu et al. have shown that defect datasets such as Defects4J suffer from reproduction failures (i.e., reported bugs become non-reproducible) as time passes since their creation. However, it remains unclear whether these findings generalize to quantum software defect datasets. We therefore conduct a replication study of the prior work using Bugs4Q, a widely used dataset of real-world bugs in quantum programs. Our analysis includes 77,700 quantum program executions of 37 Bugs4Q artifacts across 21 core-library versions. The experimental results showed that the reproducibility of Bugs4Q dropped from 62.2% on Qiskit v0.20.1 to 16.2% on v2.3.1, the latest version as of April 1, 2026. A manual inspection of the root causes further indicated that 93.6% of the failures were dependency-related. While these findings are consistent with those of the prior work, we also observed differences. In particular, most reproduction failures in Bugs4Q cannot be resolved merely by adjusting dependency versions; instead, they require source-code modifications such as migrating import paths and API invocations. Based on this observation, we curated Bugs4Q-Robust, a patched version of Bugs4Q to restore reproducibility. Bugs4Q-Robust increases reproducibility from 16.2% to 78.4% on Qiskit v2.3.1. Our findings highlight the importance of continuous dataset maintenance in the rapidly evolving quantum software ecosystem.
Large Language Models (LLMs) are increasingly used for code generation, yet quantum code generation is still evaluated mostly within single frameworks, making it difficult to separate quantum reasoning from framework familiarity. We introduce QuanBench+, a unified benchmark spanning Qiskit, PennyLane, and Cirq, with 42 aligned tasks covering quantum algorithms, gate decomposition, and state preparation.
We evaluate models with executable functional tests, report Pass@1 and Pass@5, and use KL-divergence-based acceptance for probabilistic outputs. We additionally study Pass@1 after feedback-based repair, where a model may revise code after a runtime error or wrong answer. Across frameworks, the strongest one-shot scores reach 59.5% in Qiskit, 54.8% in Cirq, and 42.9% in PennyLane; with feedback-based repair, the best scores rise to 83.3%, 76.2%, and 66.7%, respectively. These results show clear progress, but also that reliable multi-framework quantum code generation remains unsolved and still depends strongly on framework-specific knowledge.