In vLLM v0.23.0, `vllm:inter_token_latency_seconds` use it to inspect token-level decode cadence; do not equate its histogram count with requests or its mean with request-weighted `TPOT`.

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

PrometheusStatLoggerV1 metrics

vLLM v0.23.0 v1 metrics stats.py(vllm/v1/metrics/loggers.py:360-405)

supportsprimarysource_codecode_introspection

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

In vLLM v0.23.0, vllm:inter_token_latency_seconds use it to inspect token-level decode cadence; do not equate its histogram count with requests or its mean with request-weighted TPOT.