Promptfoo

Verified

Open-source, CLI-first prompt testing and LLM evals.

Promptfoo is an open-source (MIT) tool for evaluating prompts and models with declarative test cases, assertions, and red-teaming, designed to run locally and in CI. It is code-first and integrates into pipelines.

License

Open source (MIT)

Deployment

Self-hosted

Pricing

Open-source and free; enterprise offering available.

SDKs / languages

Any (CLI/YAML), JavaScript, Python

Founded

2023

Strengths

CI-friendly, code-first evals
Assertions + red-teaming
Free and open-source

Limitations

Less of a collaborative UI
Config-driven learning curve

Also compared in

LLM Observability & Evals

Alternatives

Agenta

Open-source LLMOps: prompt playground, evals, and observability.

Braintrust

Eval-first platform for iterating on LLM quality.

Langfuse

Open-source LLM observability, evals, and prompt management.

LangSmith

Tracing, evals, and prompt management from the LangChain team.

Latitude

Open-source prompt engineering and evaluation platform.

Mirascope

Pythonic library for prompts as typed functions.

Compare all prompt engineering & management →

Facts verified 19 July 2026. Neutral summary, not an endorsement; verify current details with the vendor.