​
What is evaluations?

Evaluations help you assess the performance of your prompts. You can create custom evals to measure different dimensions of output quality.

​
Why use evaluations?

  • To measure the quality of a prompt or a model.
  • To find the best prompt or model for a specific task.
  • To optimize your prompts and models.

​
Quickstart

LLM-as-judge evaluator

Create an LLM-as-judge evaluator to measure the quality of your prompts.

Create a human evaluator

Create a human evaluator to measure the quality of your prompts.

Run experiments with testsets

Run experiments with testsets to measure the performance of your prompts.