The community member has a question about whether the FaithfulnessEvaluator and RelevancyEvaluator make new (cost) requests through the API when using Anthropic as the language model. Another community member confirms that the evaluators use LLM calls.
