Master Promptfoo with our step-by-step tutorial, detailed feature walkthrough, and expert tips.
Explore the key features that make Promptfoo powerful for ai evaluation workflows.
Promptfoo is used to test and evaluate AI applications before they reach users. The public documentation describes it as an open-source CLI and library for evaluating and red-teaming LLM apps, and the website lists products for evaluations, red teaming, guardrails, model security, MCP proxy protection, and code scanning. In practice, this means teams can compare prompts and models, test RAG factuality, look for jailbreak risks, and scan LLM application code as part of development or CI/CD.
Yes. Promptfoo’s documentation describes it as an open-source CLI and library, and the public pricing page lists a Community plan as Free Forever. The Community plan includes core evaluation and vulnerability-scanning workflows, local or self-hosted operation, all listed model providers and integrations, and red teaming up to 10k probes per month. The same pricing page also lists Enterprise and On-Premise paid options with custom pricing.
Promptfoo is more focused on systematic testing, red-teaming, and AI security checks during development, while tools such as LangSmith and Braintrust are often selected for tracing, observability, experiment tracking, or evaluation management. Promptfoo’s website lists Red Teaming, Guardrails, Model Security, MCP Proxy, Code Scanning, and Evaluations as separate product areas, which gives it a stronger security-testing orientation. Choose Promptfoo when you need adversarial testing and CI-friendly regression checks around LLM applications.
Yes, the website explicitly lists industry solutions for Financial Services, Insurance, Telecommunications, and Real Estate. It mentions examples such as FINRA-aligned security testing, policyholder data and coverage accuracy, voice and text AI agent security, and fair housing compliance testing. Those examples suggest Promptfoo is aimed at teams that need evidence-driven testing around compliance, safety, and business-specific failure modes. Teams should still validate whether the enterprise deployment, audit, and contract terms meet their own regulatory requirements.
The website presents both evaluation and protection-oriented products. Evaluations cover prompt, model, and RAG testing, while Guardrails are described as real-time protection against jailbreaks and adversarial attacks. The site also lists an MCP Proxy for securing Model Context Protocol communications and Code Scanning for finding LLM vulnerabilities in IDE and CI/CD. That combination means Promptfoo can support pre-deployment testing and some runtime protection use cases, although production observability may still require a separate tracing or monitoring tool.
Promptfoo’s public pricing page lists Community as Free Forever at $0/month, Enterprise as Custom, and On-Premise as Custom. Community includes all LLM evaluation features, all model providers and integrations, red teaming up to 10k probes per month, local or self-hosted operation, vulnerability scanning, and community support. Enterprise and On-Premise do not publish exact monthly or annual prices, billing periods, paid seat limits, minimum contract terms, standard usage caps, or automatic upgrade thresholds; teams must contact sales for a quote and final conversion terms.
Now that you know how to use Promptfoo, it's time to put this knowledge into practice.
Sign up and follow the tutorial steps
Check pros, cons, and user feedback
See how it stacks against alternatives
Follow our tutorial and master this powerful ai evaluation tool in minutes.
Tutorial updated March 2026