Schema.ai· vLLM· Schema 0.2.1
vLLM official site vLLM v0.23.0Log in

In vLLM v0.23.0, vllm:prefix_cache_hits is the token numerator for a same-label, same-window prefix-cache hit ratio.

MetricvLLM v0.23.0confidence: highverified 11 hr agoverified-docs

Metric

prefix_cache_hits

Applies when

PrometheusStatLoggerV1 metrics

Evidence

vLLM v0.23.0 metrics design docs(docs/design/metrics.md:405-432)
definesprimarydocumentationdocs_extraction

Does not establish

  • a workload-independent healthy threshold
  • semantics in any vLLM version other than v0.23.0

Guards

  • vllm_version = v0.23.0
  • metric_name = vllm:prefix_cache_hits

Freshness

Pinned release-tag semantics; do not carry to another version without continuity evidence.