Vibe train SLM evals in Claude Code
Build production-ready AI judges without leaving Claude
Start Zero shot or with CSV
Get your endpoint in minutes
> 15% + accuracy improvement
< 100ms accuracy and low cost
LLM OUTPUT
"Here's a clear explanation..."
P
Eval
Helpful
LLM OUTPUT
"Here's a clear explanation..."
P
Guardrail
Helpful
LLM OUTPUT
"Here's a clear explanation..."
P
Classifier
Helpful