Skip to main content
Foglamp prices every span at ingest, the moment it arrives, from the token dimensions the provider reported. Cost is never recomputed downstream, so what you see in a trace is what was charged when it landed.

​
How pricing works

Each usage dimension is priced independently against a model price table sourced from OpenRouter, refreshed every 24 hours. The dimensions priced are prompt, completion, request, image, web-search, internal reasoning, cache-read, and cache-write tokens. The span detail panel shows a pricing source so you can see where each cost came from.
A model Foglamp can’t find a price for is labelled (unknown) in cost breakdowns — never silently treated as $0. The Overview KPI’s ”% priced” tells you what fraction of LLM spans were priced; a low number usually means you’re running a model OpenRouter doesn’t list yet.

​
Where cost shows up

  • Overview — total cost, cost-over-time (stacked by model), and per-model / per-agent cost tables.
  • Traces — per-span cost and pricing source.
  • Agents and Workflows — cost rolled up by actor and by run.
  • Alerts — alert on cost over a window.