To systematically evaluate, compare, and red-team LLM applications to ensure their security and reliability.
Promptfoo provides a CLI and library for rigorously evaluating and red-teaming LLM applications. It lets developers compare different LLM models and test prompts using declarative configuration files, integrating with CI/CD pipelines. Users can quickly get started by installing it via `npm` and running `promptfoo eval` and `promptfoo view` to see results.
To systematically evaluate, compare, and red-team LLM applications to ensure their security and reliability.
Developers and MLOps engineers building LLM applications who need to rigorously test, compare, and secure their AI systems.