Opik

Opik is an open-source platform from Comet that helps teams build and ship large language model applications with confidence. It provides a unified workspace for logging prompts, tracking model outputs, and running evaluations across different LLM providers. The tool captures every interaction so you can debug issues, compare model versions, and catch regressions before they reach production.

Opik integrates with popular frameworks like LangChain, LlamaIndex, and OpenAI, and supports Python and REST APIs for custom instrumentation. Its evaluation engine lets you define scoring functions, run batch tests, and visualize results in a dashboard. The platform also offers A/B comparison views, trace-level debugging, and built-in support for common LLM metrics like latency, token usage, and response quality.

The project is fully open source under the Apache 2.0 license and can be self-hosted or used with Comet's cloud offering. It is designed for ML engineers and developers who need a practical way to validate LLM behavior without locking into a proprietary ecosystem. Opik is actively maintained on GitHub with regular releases.