Skip to content

Qrisp Plugin Getting Started

Install

pip install "openquantum-sdk[qrisp]"

This installs the Qrisp backend along with qrisp, openquantum-sdk (Core SDK), and openquantum-sdk-qiskit.

Create a Backend

The OpenQuantumBackend builds an OpenQuantumService with from_env(). Authentication is resolved in this order: a pre-built service, then a token, then client_id + client_secret (or an SDK-key file at OPENQUANTUM_SDK_KEY). Each explicit value falls back to the matching environment variable. Omitted service URLs also read OPENQUANTUM_SCHEDULER_URL, OPENQUANTUM_MANAGEMENT_URL, OPENQUANTUM_KEYCLOAK_BASE, and OPENQUANTUM_KEYCLOAK_REALM. If nothing resolves, construction raises AuthenticationError.

export OPENQUANTUM_CLIENT_ID="s_your_client_id"
export OPENQUANTUM_CLIENT_SECRET="your_client_secret"

from qrisp_openquantum import OpenQuantumBackend

oq = OpenQuantumBackend(backend="rigetti:cepheus-1-108q", shots=1024)

With Explicit Credentials

oq = OpenQuantumBackend(
    backend="rigetti:cepheus-1-108q",
    shots=1024,
    client_id="s_your_client_id",
    client_secret="your_client_secret",
)

With a Pre-built Service

Reuse an existing OpenQuantumService (for saved accounts, custom URLs, or keyring-backed credentials):

from openquantum_sdk_qiskit import OpenQuantumService
from qrisp_openquantum import OpenQuantumBackend

service = OpenQuantumService.from_saved_account(name="default")
oq = OpenQuantumBackend(backend="rigetti:cepheus-1-108q", shots=1024, service=service)

Run a Program

Pass the backend to get_measurement on a QuantumVariable:

from qrisp import QuantumVariable, h, cx

qv = QuantumVariable(2)
h(qv[0])
cx(qv[0], qv[1])

print(qv.get_measurement(backend=oq, shots=1024))
# {'00': 0.51, '11': 0.49}

Note

get_measurement returns Qrisp's decoded outcomes as probabilities, not shot counts. This is Qrisp's own normalization and applies to every backend. To get integer counts, run a low-level QuantumCircuit with QuantumCircuit.run(backend=oq, shots=...), shown next.

Or run a raw QuantumCircuit directly, which returns integer counts:

from qrisp import QuantumCircuit

qc = QuantumCircuit(2)
qc.h(0)
qc.cx(0, 1)
qc.measure([0, 1])

counts = qc.run(backend=oq, shots=1024)
# e.g. {'00': 518, '11': 506}

Backend-Specific Configuration

Some backends require extra parameters. Simulators in the Quantum Rings family, for example, require an engine. Pass these through configuration_data:

oq = OpenQuantumBackend(
    backend="qr:gpu-gb10",
    shots=1024,
    configuration_data={"engine": "amber_quantum_rings"},
)

Note

Only send the fields a backend actually needs. Conditional schema fields (for example a threshold that only applies when performance="CUSTOM") should be set explicitly and only when they apply.

Cost and Scheduling Controls

oq = OpenQuantumBackend(
    backend="ionq:forte-1",
    shots=1024,
    job_subcategory_id="ml:qcl",   # classify your workload
    execution_plan="private",       # "auto" | "public" | "private"
    queue_priority="priority",      # "auto" | "standard" | "priority" | "instant"
    organization_id="your-org-uuid",
    job_timeout_seconds=3600,       # max wait (default 86400 / 1 day)
)

Next Steps