DeepEval is the open-source LLM evaluation framework for testing and benchmarking LLM applications — 50+ plug-and-play metrics for AI agents, RAG, chatbots, and more.