Kepler Q-Max — Sonicium Quantum Lab

Developers · API Reference

One endpoint. Quantum on tap.

REST + JSON over HTTPS. Bearer-token auth. Idempotency-Key support on every write. SDKs for Python, TypeScript, and Go.

Base URL

https://api.keplerqmax.sonicium.ltd

All requests require Authorization: Bearer sk_live_... and Content-Type: application/json.

Endpoints

POST/v1/predictRun hybrid quantum-classical inference for any pre-trained Kepler task.
POST/v1/classifyVariational quantum classifier — returns label + confidence.
POST/v1/optimizeQAOA optimizer for routing, scheduling, and portfolio problems.
POST/v1/jobsSubmit long-running QPU jobs; poll or subscribe via webhook.
GET/v1/jobs/{id}Fetch status, result, and quantum metadata for a single job.

Example: classify

POST /v1/classify
{
  "model": "kepler-q-max",
  "task": "target_recognition",
  "features": [[0.12, 0.84, 1.7, ...]]
}

200 OK
{
  "results": [
    { "label": "Hostile", "confidence": 0.91, "range_m": 1840 }
  ],
  "model_version": "qmax-25q-2026.05",
  "latency_ms": 84
}

Errors

400invalid_requestMalformed payload or unknown parameter.
401unauthorizedMissing or invalid API key.
403forbiddenKey lacks required scope.
404not_foundResource does not exist.
409conflictConcurrent modification — retry with latest version.
429rate_limitedPlan or per-key throttle hit. Use Retry-After.
500internalUnexpected error. Safe to retry with idempotency-key.

Rate limits

Free: 60 req/min. Pro: 600 req/min. Enterprise: negotiated. Limits are per-key and surface as X-RateLimit-Remaining / X-RateLimit-Reset headers.