{
  "apiVersion": "catalog.swiss/v2",
  "site": "https://modelsphere.github.io/mock-llmd/",
  "repo": "https://github.com/modelsphere/mock-llmd",
  "raw": "https://raw.githubusercontent.com/modelsphere/mock-llmd/",
  "commit": "48df5d51e5a399eb0ac8d27598ba69833a699c11",
  "count": 1,
  "models": [
    {
      "name": "mock-llm",
      "displayName": "Mock LLM",
      "description": "A CPU-only stand-in for an LLM server, for tests. mock-llmd answers /v1/chat/completions, /v1/completions and /v1/messages, streaming or not, with synthetic tokens at a TTFT and chunk rate set per request. No GPU and no weights, so routing, scaling and deploy flows can run on any node.",
      "family": "mock",
      "tags": [
        "chat",
        "test"
      ],
      "defaultVariant": "mock-cpu",
      "reports": [],
      "tuning": [],
      "variants": [
        {
          "id": "mock-cpu",
          "version": "1.0.0",
          "releaseDate": "2026-11-05",
          "commit": "48df5d51e5a399eb0ac8d27598ba69833a699c11",
          "path": "models/mock-llm/mock-cpu.yaml",
          "digest": "sha256:e53524e11262d99c96fe93e29d5da5ad8f6225bda36d8c9a1f8642eb852c1712",
          "history": "models/mock-llm/mock-cpu.history.json",
          "servedName": "mock-llm",
          "engine": "mock",
          "description": "mock-llmd on CPU: no GPU, no weights. For tests only -- the responses are synthetic.",
          "link": "https://github.com/modelsphere/mock-llmd",
          "chart": {
            "name": "vllm",
            "version": ">=0.7.6"
          },
          "image": {
            "repository": "ghcr.io/modelsphere/mock-llmd",
            "tag": "0.1.0"
          },
          "requires": {
            "gpus": 0,
            "topology": "single-node"
          }
        }
      ]
    }
  ]
}