Skip to main content
POST
/
api
/
public
/
v2
/
traces
/
analytics
/
custom-analytics
curl --request POST \
  --url https://api.promptlayer.com/api/public/v2/traces/analytics/custom-analytics \
  --header 'Content-Type: application/json' \
  --header 'X-API-KEY: <api-key>' \
  --data '
{
  "filter_group": {
    "logic": "AND",
    "filters": [
      {
        "field": "trace_name",
        "operator": "is",
        "value": "agent-turn"
      },
      {
        "field": "trace_start",
        "operator": "between",
        "value": [
          "2026-07-01T00:00:00Z",
          "2026-07-08T00:00:00Z"
        ]
      }
    ]
  },
  "customCharts": [
    {
      "id": "slowest_tools",
      "title": "Slowest Tools",
      "chartType": "bar",
      "metric": "max",
      "metricField": "span_duration_ms",
      "groupByField": "span_tool_name"
    }
  ]
}
'
{
  "success": true,
  "customCharts": [
    {
      "id": "<string>",
      "title": "<string>",
      "series": [
        {
          "key": "<string>",
          "label": "<string>"
        }
      ],
      "data": [
        {}
      ],
      "derivedInsights": [
        {
          "type": "ratio",
          "label": "<string>",
          "numeratorSeriesKey": "<string>",
          "denominatorSeriesKey": "<string>",
          "value": 123
        }
      ]
    }
  ]
}
Run custom aggregations over your traces. Same chart engine as Request Analytics — Custom Queries, but each query aggregates at one of two levels:
  • Trace-level (trace_* fields) — measure whole traces: durations, cost, tokens, span counts.
  • Span-level (span_* fields) — measure the individual spans inside matching traces: e.g. the slowest tools, tool call counts, span duration distributions.
filter_group selects which traces participate (by trace_name, a trace_start range, span conditions, etc.); span-level charts then aggregate across every span of those traces.

​
Metrics

count, sum, avg, min, max, percentile (with percentile 0–100).

​
Metric fields (metricField)

LevelFields
Tracetrace_duration_ms, trace_total_cost_usd, trace_total_tokens, trace_input_tokens, trace_output_tokens, trace_span_count, trace_depth
Spanspan_duration_ms, span_cost_usd, span_tokens, span_input_tokens, span_output_tokens
Duration values are returned in seconds (converted from milliseconds internally).

​
Group-by fields (groupByField)

LevelFields
Tracetrace_status, trace_name, trace_models_used, trace_prompt_ids, trace_workflow_ids, trace_tool_names
Spanspan_tool_name, span_name, span_type, span_kind, span_status

​
Constraints

  • A chart must stay at one level: span-level fields cannot be mixed with trace-level fields in the same chart.
  • Span-level charts do not support timeSeries or metadata-key breakdowns.
  • Chart shapes beyond bar/line/area are supported: pie, donut, histogram (with histogramField), heatmap (with secondaryGroupByField), treemap/sunburst (with hierarchyFields).

​
Response shape

Identical to Request Analytics — Custom Queries: each entry in customCharts has id, series descriptors ({ key, label, unit }), and data rows. Count series are labeled “Traces” for trace-level charts and “Spans” for span-level charts.

Authorizations

​
X-API-KEY
string
header
required

Body

application/json
​
customCharts
object[]
required
Required array length: 1 - 10 elements
​
filter_group
object

Nested filter tree. AND/OR are cross-span; SPAN_AND/SPAN_OR require one and the same span to satisfy every/any branch (span-level fields only).

Response

Custom chart results in the order requested.

​
success
enum<boolean>
required
Available options:
true
​
customCharts
CustomAnalyticsResult · object[]
required

Results in the same order as the input customCharts array.