You can easily test different prompt and model combinations. Variable parametrization is also supported.
Results will show the LLM response, latency, cost and optionally the Evaluator scores:
Last updated 5 months ago
