Skip to main content
The Request Logs page shows every request that has passed through your gateway in real time. You can search, filter, and inspect the full details of any individual call.

​
Reading the Logs Table

Each row in the table represents a single gateway request and shows:
  • Request ID — A unique identifier in the format req_xxxxxxxx
  • Timestamp — When the request was received
  • Model — The resolved model (e.g. gpt-4o, claude-opus-4-5)
  • Provider — Which upstream provider handled it
  • Tokens — Prompt + completion token breakdown
  • Cost — Exact cost in USD
  • Latency — End-to-end duration in milliseconds
  • Cache — HIT or MISS
  • Status — HTTP status from the upstream

​
Filtering

Use the filter bar at the top of the page to narrow results by:
  • Date range — From/To timestamps
  • Provider — e.g. openai, anthropic, google
  • Model name
  • Status code — e.g. 200, 429, 500
  • API Key — See which key generated traffic
  • Custom properties — Filter by any metrix-property-* tag you sent

​
Inspecting a Single Request

Click any row to open the request detail panel. You will see:
  • The full prompt text sent to the model
  • The full response text received
  • Whether a fallback was triggered and the original vs final provider
  • The retry count if retries were attempted
  • Session trace information if metrix-session-id was provided
Prompt and response bodies are only stored when Data Retention is enabled in your workspace settings.