{
  "name": "pinecone-S-latency",
  "description": "Pinecone SERVERLESS (kein Pod-Quota -> keine s1.x1-Paritaet, Fallback; Latenz client+Netz, nur Recall vergleichbar). Stufe S, Single-Query.",
  "db": "pinecone",
  "stufe": "S",
  "dim": 1024,
  "variant": "A",
  "workload": "topk",
  "index": {
    "type": "serverless",
    "params": {
      "spec": "serverless",
      "region": "aws-us-east-1",
      "metric": "cosine"
    }
  },
  "queries": {
    "n": 10000,
    "concurrency": 1
  },
  "index_name": "bench-s-latency",
  "pre_run_reset": false
}