qdevops-io 0.1.1

Official Python SDK for the QDevOps quantum reproducibility platform (qdevops.io).

qdevops-io — Python SDK

Official Python client for the QDevOps quantum reproducibility platform.

The distribution name on PyPI is qdevops-io; the import path is qdevops (same convention as python-dateutil → dateutil).

• Submit Bell / GHZ / QFT / Grover / VQE circuits to the simulator, IBM Quantum, or AWS Braket
• Poll runs to completion, fetch result payloads, diff two runs
• Export pinned environment lockfiles
• Publish results to the public benchmark dashboard at qdevops.io/bench

Install

pip install qdevops-io

Requires Python 3.10+.

Authenticate

Mint a personal access token at qdevops.io/account/tokens, then either:

export QDEVOPS_API_TOKEN="qass_pat_..."

…or pass it explicitly:

from qdevops import Client

qd = Client(api_token="qass_pat_...")

Quickstart — run a Bell state on the simulator

from qdevops import Client

with Client() as qd:
    run = qd.submit_run(
        project_id=42,
        circuit="bell",
        backend="simulator",
        params={"shots": 2048},
    )
    print("queued", run.run_id, run.circuit_hash)

    full = qd.wait_for_run(run.run_id, timeout=120)
    if full.status == "succeeded":
        print("counts:", full.result.get("counts"))
    else:
        print("failed:", full.failure_reason)

Compare backends

for backend in ("simulator", "ibm", "braket"):
    run = qd.submit_run(project_id=42, circuit="bell", backend=backend)
    finished = qd.wait_for_run(run.run_id)
    print(backend, finished.result.get("expectation"))

Publish a benchmark result

Requires a token with the benchmarks:write scope.

qd.publish_benchmark(
    circuit="bell",
    backend="simulator",
    commit_sha="abc1234",
    fidelity=0.984,
    shots=2048,
    duration_ms=210,
    sdk_version="0.1.1",
)

Errors

All non-2xx responses raise a subclass of qdevops.APIError:

Exception	HTTP
AuthenticationError	401
ForbiddenError	403
NotFoundError	404
ConflictError	409
APIError (generic)	4xx/5xx

Client.wait_for_run raises qdevops.errors.TimeoutError if a run does not reach a terminal state within the requested timeout.

Reference

See the OpenAPI spec at api.qdevops.io/docs.

License

Proprietary © QDevOps.io