Triton cumulative compute-infer duration
triton-inference-server · counter
Fields
id
metric-exposures/w3m-triton-inference-server-nv-inference-compute-infer-duration-usMetric name
nv_inference_compute_infer_duration_us
Metric type
counter
Unit
microseconds
Exposure
prometheus
Endpoint
:8002/metrics
Labels
modelversion
Enabled by default
yes
notes
Triton main, docs/user_guide/metrics.md, read 2026-10-04. On by default: 'By default, these metrics are available at http://localhost:8002/metrics', and the endpoint is pure pull - 'not pushed or published to any remote server'. tritonserver --allow-metrics=false turns the whole thing off. Labels are model and version, and crucially the LATENCY families are per MODEL, per request-update - not per-request observations as in vLLM/SGLang/TGI. Framework-backend only.
Notes
Triton main, docs/user_guide/metrics.md, read 2026-10-04. On by default: 'By default, these metrics are available at http://localhost:8002/metrics', and the endpoint is pure pull - 'not pushed or published to any remote server'. tritonserver --allow-metrics=false turns the whole thing off. Labels are model and version, and crucially the LATENCY families are per MODEL, per request-update - not per-request observations as in vLLM/SGLang/TGI. Framework-backend only.