Metric & Benchmark
P95 Latency
95th-percentile request latency — 5% of requests are slower than this.
Definition
P95 latency captures the slow tail that defines user-perceived performance. A high P95 reveals overload, GC pauses or cold-start issues invisible to the median. Many SLOs are set on P95 or P99 rather than P50.
Common use cases
- SLO definition
- Performance budgets
- Alerting