Tell us what you want to compute. We'll tell you where to run it.
Qlro is a vendor-neutral quantum device recommender. Give it a workload — a Qiskit circuit or one of 60 named industry workload templates — and it ranks every available quantum device by how well that device fits your specific workload, grounded in real benchmark data from Metriq, not vendor marketing.
import qlro
# Domain-vocabulary entry point — no quantum-circuit knowledge required
result = qlro.recommend_intake(
template_id="industry.finance.option_pricing_qae",
accuracy_tier="high_precision",
data_scale="typical",
total_budget_usd=10_000,
)
print(result["primary"]) # → 'H2-2'
print(result["reasons"]) # → templated, deterministic explanationpip install qlroApache 2.0. Python 3.11+. Ships with a snapshot of the Metriq benchmark dataset and 60 procurement-ready workload templates across pharma, finance, and manufacturing.
- R&D leads / procurement officers who need to pick a quantum vendor without having a quantum specialist on staff.
- Government R&D evaluators who need vendor-neutral, citable evidence for grant or procurement filings.
- Quantum engineers who already have a circuit and want to compare devices on its specific physics.
- Live dashboard — qlro.io/dashboard — daily-views control center across 13 devices.
- Industry workloads — qlro.io/workloads — 49 named procurement-ready templates.
- Public accuracy log — qlro.io/accuracy — community-submitted (predicted, observed) fidelity pairs with monthly DOI-stamped snapshots.
- Browser simulator — qlro.io/simulator — 5-minute interactive walkthrough.
Qlro is the reference implementation of WCPP (Workload-Conditioned Physical Projection) — a vendor-neutral framework that maps real benchmark data (Metriq) onto four physics-grounded capability axes (Γ connectivity, Φ coherence, F fidelity, T throughput) and then composes a workload-specific fit score. A Circuit Survival Estimator (CSE) on top predicts output fidelity end-to-end. Adaptive 1-shot calibration recovers cross-vendor RMSE 82–94% from cheap calibration circuits. The full math, axioms, and proofs are in the WCPP paper (DOI below).
Every quantum-circuit execution can flow into the public accuracy dashboard automatically — two lines of code:
# AWS Braket
import qlro.autolog.braket as qlbraket
qlbraket.enable()
# Qiskit
import qlro.autolog.qiskit as qlqiskit
backend = qlqiskit.wrap(your_backend)After that, every task.result() or job.result() call posts an anonymous (predicted, observed) pair. No PII, no API keys, no manual log_outcome() plumbing.
qlro workload --list --industry pharma
qlro workload industry.finance.option_pricing_qae --params '{"num_state_qubits":4}'
qlro recommend my_circuit.qasm --category chemistry --all
qlro doctor iqm_garnet # snapshot freshness + drift check
qlro calibrate iqm_garnet # 1-shot adaptive calibrationqlro --help lists every subcommand.
The WCPP framework is published on Zenodo with a permanent DOI:
DOI: 10.5281/zenodo.19785800 (v1.2, post-reviewer round-3)
@misc{oh2026wcpp,
author = {Oh, Yeonwoo},
title = {{Workload-Conditioned Physical Projection: A Vendor-Neutral
Framework for Quantum Device Selection}},
year = {2026},
publisher = {Zenodo},
version = {1.2},
doi = {10.5281/zenodo.19785800},
url = {https://doi.org/10.5281/zenodo.19785800}
}Concept DOI (always resolves to latest): 10.5281/zenodo.19601378.
The Python package is published on PyPI. Install via pip install qlro to get the full implementation, including the 60 workload templates and the CSE forward model.
This GitHub repository hosts the project metadata, license, and issue tracker. Several algorithmic implementations (1-shot adaptive calibration, the cryptographically-bound decision-record minting flow) are subject to pending patent applications and are distributed only via the published PyPI wheel; see NOTICE.md for the current open-source / patent status.
Bug reports, questions, and feature requests: github.com/linsletoh/qlro/issues.
Apache 2.0 — see LICENSE.