PrismHubUNDERSTAND THE FULL REQUESTClarity at every step.
Clarity at every step.
From input to outcome.
A response is only part of the story. Connect model choice, request attempts, and token usage in one traceable view.
REQUEST EXPLORERSample dataset · not live traffic
Requests in sample6
Total sample tokens15,976
Fallback recovered1 request
Model / taskOutcomeTokensLatency
Showing 6 sample requests. Select a row to inspect its trace.
Follow the attempts
Separate a successful first response from a request recovered by a fallback. Inspect the status that triggered the next step.
Understand usage
Track input and output tokens independently. Connect the reported usage to the rate rules for the model you selected.
Keep context useful
Use request IDs to debug across your stack. Avoid logging secrets or unnecessary prompt content in application traces.