Cirq

{
  "service_type": "platform",
  "base_url": "https://quantumai.google/cirq",
  "auth_method": "none",
  "auth_config": {},
  "endpoints": [],
  "pricing_model": {
    "type": "free",
    "details": {
      "note": "Cirq is an open-source Python framework released under the Apache 2.0 license. It is free to use, modify, and distribute. Some execution backends (e.g., Google Quantum AI hardware, cloud quantum services) may have separate access or pricing via their respective providers."
    }
  },
  "rate_limits": {},
  "capabilities": [
    "Quantum circuit construction and editing (qubits, gates, moments)",
    "Noisy Intermediate-Scale Quantum (NISQ) circuit simulation",
    "Circuit optimization and compilation",
    "Execution on quantum simulators (local and cloud)",
    "Execution on real quantum hardware via providers (e.g., Google Quantum AI, IonQ, AQT, Pasqal)",
    "Integration with quantum libraries and tools through interoperability",
    "Hardware-agnostic gate and operation abstractions",
    "Parameterized circuits and sweeps",
    "OpenFermion and TensorFlow Quantum integrations",
    "Open-source, extensible Python API"
  ],
  "raw_analysis": "Cirq is an open-source Python framework developed by Google Quantum AI for building, manipulating, simulating, and executing quantum circuits, with a focus on NISQ-era devices. It is aimed primarily at quantum computing researchers, developers, and educators who need a programmatic way to define circuits and run them on simulators or real quantum hardware.\n\nMaturity and ecosystem: Cirq is one of the more mature quantum programming frameworks, with active development, extensive documentation, tutorials, and a sizable community. It serves as the foundation for higher-level tools including TensorFlow Quantum, and it interoperates with OpenFermion and other quantum software.\n\nIntegrations and execution: Cirq itself is a library, not a hosted REST API platform, so it has no public HTTP endpoints or authentication layer. Execution on real hardware typically happens through provider-specific packages (e.g., cirq-google for Google Quantum AI, cirq-ionq, cirq-aqt, cirq-pasqal, cirq-rigetti) and cloud services, each with its own access model. Local simulation is unrestricted and requires no API keys.\n\nPricing: The core framework is free and open source under Apache 2.0. Any costs arise only from third-party quantum cloud/hardware backends.\n\nLimitations for integration: Because there is no standardized public REST API, automated integration must be done through the Python package interface rather than HTTP. Rate limits, quotas, and authentication are dictated by whichever hardware or cloud provider is used, not by Cirq itself."
}

1/3 tests passed

{
  "overall": 72,
  "dimensions": {
    "token_efficiency": 9.0,
    "first_try_success": 8.5,
    "response_parseability": 7.5,
    "error_clarity": 7.0,
    "doc_quality": 8.5,
    "auth_simplicity": 9.5,
    "latency": 9.0,
    "consistency": 8.0
  },
  "pricing_normalized": {
    "model": "free_open_source",
    "license": "Apache 2.0",
    "cost": 0,
    "backend_costs": "varies_by_provider",
    "note": "Framework is free; real hardware execution may incur provider-specific costs"
  },
  "issues": [
    "No robots.txt (404) — crawlers/agents lack guidance",
    "No llms.txt (404) — no agent-optimized documentation entry point",
    "Website returned HTML shell with Google Sign-In meta tags rather than structured API docs at root",
    "Hardware execution requires separate provider accounts with independent auth flows"
  ],
  "recommendations": [
    "Publish robots.txt and llms.txt for agent-friendly discovery",
    "Add a machine-readable docs endpoint (e.g., /api or OpenAPI spec) describing Cirq's public classes",
    "Provide a stable JSON manifest of capabilities, install command, and version pinning",
    "Document provider auth requirements clearly so agents can pre-flight credential checks"
  ]
}