Quick navigation
Products
Each product has a standalone REST API, SDK, and CLI. All products expose MCP tools for agent integration.
The operating system for quantum hardware. REST API, Python SDK, CLI, and MCP server. Runs circuits across IBM, IonQ, AWS Braket, and on-prem QPUs from a single endpoint.
Quantum-resistant key management and hardware-backed licensing. REST API, Python SDK, and CLI for enterprise key lifecycle, revocation, and audit.
Softquantus AI server for quantum-classical model training, inference, and reranking. Integrates with QCOS for quantum feature extraction in LLM pipelines.
QCOS MCP Server
The QCOS MCP (Model Context Protocol) server exposes all Softquantus services as structured tools that any MCP-compatible AI agent can call — Claude, LangChain, AutoGen, CrewAI, and others.
{
"mcpServers": {
"softquantus": {
"command": "uvx",
"args": ["softquantus-mcp"],
"env": {
"QCOS_API_KEY": "your-api-key",
"QCOS_BASE_URL": "https://api.softquantus.com"
}
}
}
}| Tool | Description | Parameters |
|---|---|---|
qcos_submit_circuit | Submit a quantum circuit (OpenQASM 3 or Qiskit) to a backend and return job ID. | circuit: str, backend: str, shots?: int |
qcos_get_job_result | Retrieve results and DRR reliability score for a completed job. | job_id: str |
qcos_list_backends | List available quantum backends with real-time calibration status. | — |
qcos_run_vqe | Execute a VQE workflow for molecular simulation on a target backend. | molecule: str, backend: str, ansatz?: str |
qcos_run_qaoa | Execute a QAOA workflow for combinatorial optimization. | problem: object, backend: str, depth?: int |
quantumlock_generate_key | Generate a post-quantum key pair using CRYSTALS-Kyber or ML-KEM. | algorithm: str, label?: str |
quantumlock_verify_license | Verify a hardware-bound license artifact and return chain-of-trust proof. | artifact_id: str, device_fingerprint: str |
synapsex_rerank | Rerank a set of documents using quantum-enhanced similarity scoring. | query: str, documents: list[str], top_k?: int |
Agent & external service integration
Softquantus services are designed to be consumed by AI agents and connected to external cloud providers, LLM services, and HPC clusters.
- Azure Quantum
- AWS Braket
- IBM Quantum
- Google Quantum AI
- Anthropic Claude
- OpenAI GPT
- Mistral
- Llama (self-hosted)
- Claude Agents (MCP)
- LangChain Tools
- AutoGen
- CrewAI
- LUMI Supercomputer
- EuroHPC
- NVIDIA CUDA
- Custom GPU clusters
API overview
All services expose a versioned REST API with OpenAPI specs, bearer token auth, and webhook support.
QCOS: https://api.softquantus.com/v2 QuantumLock: https://lock.softquantus.com/v2 SynapseX: https://synapsex.softquantus.com/v1
# Bearer token (all services)
curl -H "Authorization: Bearer $QCOS_API_KEY" \
https://api.softquantus.com/v2/jobs
# Python SDK
from softqcos import QCOSClient
client = QCOSClient(api_key="sk-...")
result = client.jobs.run(circuit, backend="ibm-brisbane")# Install pip install softqcos-cli # Authenticate qcos auth login --api-key sk-... # Run a circuit qcos run circuit.qasm --backend ibm-brisbane --shots 1024 # Check job status qcos jobs get <job-id> # List backends qcos backends list --status online