Skip to main content
The Overview is the dashboard home: a single screen that answers “what is my AI app doing right now, and how does it compare to before?” Every number respects the date-range picker at the top of the page, and each KPI shows a delta against the immediately preceding window of equal length.

​
KPI cards

CardWhat it shows
Total costSummed cost for the window, with a ~$/mo extrapolation and the share of LLM spans that received a price (% priced).
Error rateErrored spans ÷ total spans.
Eval pass ratePass/fail ratio across scored traffic, with the number of checks scored (see Evals).
Latency p95p95 step latency, with p50 and TTFT p95 underneath.
RequestsSpan count, with the LLM-span subset.
TokensTotal tokens, broken into input/output.
Deltas for cost, error rate, and latency are inverted — red means the number went up (worse), green means it went down.

​
Trend charts

  • Cost over time — a stacked line chart of cost per minute, split by model. The top five models get their own series; the rest roll into Other.
  • Requests & errors — spans per minute with errors overlaid.
  • Latency — p50 / p95 / p99 per minute.
  • By model — a table of requests, tokens, p95 latency, and cost per model.
  • By agent — the same, grouped by agentName.
  • By workflow — the same, grouped by workflowName (traces with no workflow roll into Ungrouped).
Models Foglamp can’t price are labelled (unknown) in the breakdowns rather than shown as $0 — a missing price is never silently treated as free. Add a rule on the Pricing page to price them.

​
Date range

The range picker drives every view. Presets: last hour, last 24 hours (default), 7 / 30 / 90 days, today, this month, last month — plus a custom two-month calendar. Future dates are disabled.