List latency / TTFT / TPS quantiles

POST

https://api.respan.ai/api/dashboard/quantiles/

POST

/api/dashboard/quantiles/

$ curl -X POST https://api.respan.ai/api/dashboard/quantiles/ \
>      -H "Authorization: Bearer <respanApiKey>" \
>      -H "Content-Type: application/json" \
>      -d '{}'

200Successful

1 [
2   {
3     "date_group": "2024-04-20T14:00:00Z",
4     "latency_p_50": 0.85,
5     "latency_p_90": 1.2,
6     "latency_p_95": 1.5,
7     "latency_p_99": 2.3,
8     "ttft_p_50": 0.4,
9     "ttft_p_90": 0.7,
10     "ttft_p_95": 0.9,
11     "ttft_p_99": 1.4,
12     "tps_p_50": 15,
13     "tps_p_90": 22.5,
14     "tps_p_95": 25,
15     "tps_p_99": 30
16   },
17   {
18     "date_group": "2024-04-20T15:00:00Z",
19     "latency_p_50": 0.9,
20     "latency_p_90": 1.3,
21     "latency_p_95": 1.6,
22     "latency_p_99": 2.5,
23     "ttft_p_50": 0.45,
24     "ttft_p_90": 0.75,
25     "ttft_p_95": 1,
26     "ttft_p_99": 1.5,
27     "tps_p_50": 16,
28     "tps_p_90": 23,
29     "tps_p_95": 26,
30     "tps_p_99": 31
31   }
32 ]

Returns p50, p90, p95, and p99 distributions for latency, time-to-first-token, and tokens-per-second, bucketed by time_tick.

Authentication

AuthorizationBearer

Use your Respan API key for Respan API authentication. Enter only the Respan API key value; clients send Authorization: Bearer <RESPAN_API_KEY>. For /api/responses, OpenAI or Azure OpenAI provider credentials go in Settings -> Providers or the request body credential_override field, not in this auth field.

Query parameters

summary_typeenumOptionalDefaults to all

Preset time range. Use this or explicit start_time / end_time.

datedateOptional

Base date used with summary_type presets.

start_timedatetimeOptional

Optional explicit ISO start time.

end_timedatetimeOptional

Optional explicit ISO end time.

time_tickenumOptionalDefaults to hour

Bucket granularity for time-series responses.

Allowed values:

timezone_offsetdoubleOptionalDefaults to 0

Timezone offset, in hours, used when resolving preset ranges.

Request

This endpoint expects an object.

filtersobjectOptional

Filter criteria. See Filters API Reference for operator syntax.

metrics_to_aggregatelist of enumsOptional

Metric subset for quantile aggregation.

Response

Successful response.

date_groupdatetime

Bucket start time, ISO 8601.

latency_p_50double

latency_p_90double

latency_p_95double

latency_p_99double

ttft_p_50double

ttft_p_90double

ttft_p_95double

ttft_p_99double

tps_p_50double

tps_p_90double

tps_p_95double

tps_p_99double