This page explains how content is created and reviewed at ArtPrompts Generator, who is responsible for it, and how we handle mistakes. We publish free guides and prompt references, and we want you to know exactly how much you can trust them — and where the honest limits are.

Who writes and reviews our content

Every guide is written or reviewed by a named member of our team: Elena Vasquez (Founder & Creative Director), Kai Nakamura (Lead Prompt Engineer) or Noah Bennett (Staff Writer). We do not publish anonymous articles. Technical claims — anything about parameters, model behaviour or settings — are checked by Kai. Editorial direction, clarity and tone are overseen by Elena. You can read more about each of us on our Meet the Team page.

How we make a guide

Our guides are grounded in hands-on testing rather than repackaged rumour. In practice that means:

  • We start from a real question we couldn’t answer cleanly ourselves.
  • We generate images to find the answer, usually across more than one model, and keep the prompts so results are reproducible.
  • We only recommend a technique after we have seen it work, and we describe the conditions under which it works.
  • We define our terms. If we use a word like “stylize” or “CFG,” we explain it or link to our prompt glossary.

Our full methodology lives on the How We Test page.

On accuracy and change

AI image models change constantly. A prompt or setting that produced a beautiful result on one version can behave differently after an update, and new models arrive regularly. We date-stamp or note the model versions we tested where it matters, and we try to flag when guidance is version-specific. We would rather tell you “this worked on this version and may drift” than pretend our advice is permanent.

Independence

We are an independent editorial team. Our guides reflect what we found in testing, not what any model provider would prefer us to say. We are not paid by Midjourney, Stability AI, OpenAI, Black Forest Labs or any other model maker to recommend their tools. If a technique works better on one model than another, we say so plainly. Where any content is sponsored or contains affiliate links, we disclose it — see our disclaimer for details.

Corrections

We get things wrong sometimes, and models move under our feet. When we learn that a guide is inaccurate or out of date, we correct it. Significant corrections are noted within the affected guide so you can see what changed. If you believe something we published is wrong or no longer works, please tell us through our contact page — reader reports are one of the best ways we catch drift. We aim to review and respond to correction requests promptly.

Reader trust

We do not use clickbait, we do not invent statistics, and we do not fabricate credentials or awards. We are a small team writing honestly about a fast-moving field. If we don’t know something, we say we don’t know it.