analytics

Time-to-first-token percentiles per model (streaming)

Changed on
get/v1/analytics/model-router/latency/ttft-by-model

Request

Query parameters

agent_idstring uuid nullable
operationstring nullable
limitinteger
start_datestring date required
end_datestring date required

Response

Successful Response

served_modelstring required
providerstring nullable
p50_msnumber required
p95_msnumber required
p99_msnumber required
avg_msnumber required
max_msnumber required
invocationsinteger required

Changes