Trial
Opik (by Comet) is an open-source framework combining LLM evaluation and observability: tracing, datasets, experiments, online evals and a catalog of LLM-as-judge and code metrics (hallucination, answer relevance, moderation and more), including agent-trace evaluation.
- Fully Apache 2.0 and self-hostable - the core functionality works without the commercial Comet cloud.
- Very active development and one of the fastest-growing projects in the category.