# Keryn and Promptfoo

> How Keryn compares with Promptfoo, with a source for every row.

Source: https://keryn.dev/compare/promptfoo/

[Promptfoo](https://github.com/promptfoo/promptfoo) is an open source tool for LLM evals and red teaming; it [agreed to be acquired by OpenAI](https://www.promptfoo.dev/blog/promptfoo-joining-openai/) in 2026.

Every row links the source it comes from ([how these pages are written](https://keryn.dev/compare/)). Vendor statements quote the vendor's own pages as read on 18 November 2026; products change, so follow the link before relying on a row. A missing row is not a claim that the vendor lacks something.

## What they do

| Claim | Source |
|---|---|
| "LLM evals run 100% locally - your prompts never leave your machine." | [Promptfoo repository](https://github.com/promptfoo/promptfoo) |
| Code scanning for prompt injection, PII exposure and excessive agency "uses AI agents" and runs "on Promptfoo Cloud by default". | [Promptfoo code scanning](https://www.promptfoo.dev/docs/code-scanning/) |

## Where they are stronger

| Claim | Source |
|---|---|
| Tests a live application or agent with automated red teaming; Keryn analyses code and configuration. | [Promptfoo](https://www.promptfoo.dev/) |

## Where Keryn differs

| Claim | Source |
|---|---|
| Keryn imports promptfoo configs and reaches the same verdicts: 302 configs identical, 0 differing, 115 using features Keryn does not support yet. | [promptfoo receipt](https://keryn.dev/evals/promptfoo/) |
| Keryn's code analysis is rule-based and deterministic, and runs locally. | [Scores](https://keryn.dev/docs/concepts/scores/) |

Measured results for Keryn are on the [evaluations](https://keryn.dev/evals/) page, with what each one does not show.
