{
  "schema_version": "v1",
  "name_for_human": "LLM Consensus",
  "name_for_model": "llm_consensus",
  "description_for_human": "Multi-model AI orchestration API. One call orchestrates up to 5 frontier models — GPT-5.4, Claude Opus 4.6, Gemini 3.1 Pro, Mistral Large 2, Llama 3.3 70B — into superior consensus answers.",
  "description_for_model": "LLM Consensus API orchestrates up to 5 frontier language models (GPT-5.4, Claude Opus 4.6, Gemini 3.1 Pro, Mistral Large 2, Llama 3.3 70B) to generate superior answers through iterative consensus with 5 specialist roles. Send a prompt with a mode (fast/balanced/deep) and receive a single high-quality response that outperforms any individual model. Independently benchmarked: 100% non-inferiority, 44.9% win rate, 0% loss rate across 100 expert-domain questions (financial, legal, medical, technical) with 3-judge blind evaluation. Benchmark: https://llmconsensus.io/benchmark/ Supports streaming, batch processing, conversation history, domain-specific synthesis, and completeness floor guarantees.",
  "auth": {
    "type": "service_http",
    "authorization_type": "custom",
    "custom_auth_header": "X-API-Key",
    "description": "Use X-API-Key header for orchestration, or X-Payment for x402 on-chain payments. Bearer JWT for account/billing endpoints."
  },
  "api": {
    "type": "openapi",
    "url": "https://llmconsensus.io/openapi.json",
    "is_user_authenticated": false
  },
  "logo_url": "https://llmconsensus.io/logo.png",
  "contact_email": "hello@llmconsensus.io",
  "legal_info_url": "https://llmconsensus.io/terms.html"
}