Skip to content

Promptfoo

Promptfoo

CLI tool for testing and evaluating LLM prompts against test cases and RedTeam checks.

Official Site
01

Overview

Promptfoo is an LLM tooling project that helps you evaluate, observe, or optimize large language model applications. It addresses the engineering challenges of working with LLMs—non-determinism, prompt sensitivity, cost, latency, and observability—so you can ship reliable LLM features rather than brittle demos.

02

Getting Started

Install Promptfoo and integrate it with your LLM call site (wrap your model client or point it at your trace endpoint). Run an evaluation suite against your prompts to establish a baseline, then iterate on prompts and models while tracking metrics. Use the observability dashboard to inspect real production traces.

bash
# Install
pip install promptfoo

# Evaluate a prompt against a dataset
promptfoo eval --prompt ./prompts/summarize.txt --dataset ./data/eval.jsonl

# Observe production calls
promptfoo trace --serve :6006
03

Use Cases

Use cases for Promptfoo: (1) regression-testing prompts after a model upgrade; (2) A/B testing different prompt templates on real traffic; (3) tracing and debugging failed LLM calls in production; (4) cost and latency monitoring with per-prompt breakdowns; (5) automatic prompt optimization via feedback-driven search.

04

Tips

Tips for Promptfoo: version your prompts like code—store them in Git with a changelog. Build an eval set from real (anonymized) production examples, not synthetic ones. Watch for "eval hacking"—overfitting prompts to your test set. Track token cost alongside quality; a 1% quality gain at 3x cost is rarely worth it.

Ready to try Promptfoo?

Visit the official site for the latest version and full documentation.

Visit Promptfoo