Quantum instances: free, prepaid and pay-as-you-go plans

A quantum instance is how a workspace organizes access to compute, in the same style as IBM Quantum instances. Each instance defines three things:

A pricing plan — how runs submitted through it are billed.
A target allowlist — which computers/targets it may submit to (empty allowlist = any target the workspace is entitled to; up to 50 targets).
An access group — who in the workspace may use it.

Submitting a job with an instance_id makes that instance’s plan decide the billing; submitting without one uses plain prepaid billing against the workspace wallet.

The three plans

Plan	Billing	Rules
`free`	Never charges	Simulators only. One free instance per workspace. Capped at 10 minutes of runtime per month — at the cap the workload is blocked (`free_plan_limit_met`) and nothing is charged.
`prepaid` (default)	Reserve worst case, settle actual, from workspace credits	Recommended for most teams.
`paygo`	No upfront hold; usage invoiced	Released by SoftQuantus staff only after a signed usage contract. Until then everything paygo returns `403 paygo_not_released`.

Per-plan refusals you may see when running: free_plan_simulators_only (free instance pointed at real hardware), free_plan_limit_met (monthly cap reached), paygo_not_released, instance_target_not_allowed (target outside the allowlist), instance_access_denied (you are not in the access group). All are listed in Errors.

Access groups

When creating an instance you choose who can run through it:

Console label	API value	Who can use the instance
All project members	`all_members`	Every active workspace member
Owners only	`owners_only`	Workspace owners
Specific users	`custom`	Only the members you select (up to 200)

Create an instance

Console: Compute Console → Quantum instances → Create instance. Pick the plan, name it, select allowed computers and the access group. When the plan is free, real-QPU targets are not selectable and only one free instance can exist (a second returns 409 free_instance_exists). SDK / CLI / REST:

from qcoscloud import QCOSCloud
qc = QCOSCloud()

qc.create_instance(
    "research",
    plan="prepaid",                       # "free" | "prepaid" | "paygo"
    qpus=["qcos.simulator.statevector"],  # allowed targets ([] = all entitled)
    access_group="all_members",           # or "owners_only" / "custom"
)
qc.instances()

qcoscloud instances create research \
  --plan prepaid \
  --qpus qcos.simulator.statevector \
  --access-group all_members

qcoscloud instances

curl -sS -X POST https://platform.softquantus.com/api/v1/platform/quantum-instances \
  -H "Authorization: Bearer $SQ_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "research",
    "plan": "prepaid",
    "qpus": ["qcos.simulator.statevector"],
    "access_group": "all_members"
  }'

Each instance row reports its plan, allowlist, access group, the monthly runtime used against the cap (free plan), and the amount spent through it. Deleting an instance (DELETE /api/v1/platform/quantum-instances/{id}) retires it; jobs already settled keep their history.

Requesting pay-as-you-go

Pay-as-you-go is invoiced monthly with no prepaid hold, so it is gated on a contract:

Request it

In the Compute Console (or POST /api/v1/platform/quantum-instances/paygo/request with a short justification). One pending request per workspace; a second returns 409.

SoftQuantus reviews

SoftQuantus staff approve or reject the request. An approval requires a signed contract reference — a release without one is not possible, by design.

Check status, then create paygo instances

GET /api/v1/platform/quantum-instances/paygo/status returns none, pending, approved or rejected. Once approved, creating paygo instances and running through them works; before that, both return 403 paygo_not_released.

Pricing catalog

GET /api/v1/platform/quantum-instances/pricing-catalog returns the pricing source of truth: provider-published hardware prices (with source URLs) merged with the platform’s own simulator and compute rates. The console’s Quantum Computers view renders the same catalog. See Backends & pricing.

Console

Run compute

API & clients

Trust & ecosystem

Quantum instances: free, prepaid and pay-as-you-go plans

The three plans

Access groups

Create an instance

Requesting pay-as-you-go

Pricing catalog

​The three plans

​Access groups

​Create an instance

​Requesting pay-as-you-go

​Pricing catalog

The three plans

Access groups

Create an instance

Requesting pay-as-you-go

Pricing catalog