Skip to main content
API Reference: Evaluation endpoint | Python: client.eval()

client.eval()

Basic evaluation

Deep evaluation

Selective dimensions

Custom weights

Modes: basic - सिर्फ scores, 5 min cache, 1.0 credit। deep - scores + explanations + issues, 3 min cache, 3.0 credits। Weights का total 100 होना चाहिए।

Parameters

Response: EvalResult

Score labels

आगे क्या देखें

Safe Regeneration

Content को evaluate और iteratively improve करें।

Compliance

Regulatory frameworks के against content check करें।