[ LAB — LIVE · no model calls ]
Eval Result Analyzer
Paste already-scored JSON records and calculate attempts, accepted outcomes, average latency, total cost and cost per accepted result for each candidate. Failed attempts remain in the bill. The sample uses fictional Candidate A and Candidate B labels and is explicitly illustrative.
All lab toolsInteractive workbench / no model calls
How to use it
- 01
Use records with candidate, passed, costUsd and latencyMs. Each row is one evaluated attempt.
- 02
The analyzer validates types and non-negative values, groups records by candidate and computes the metrics.
- 03
Export the summary. Keep your rubric, fixture version and traces alongside it; the analyzer cannot establish whether labels are correct.