AI Image Prompts That Work: 5-Layer Text-to-Image Formula

This course contains the use of artificial intelligence.

(I am an instructor from Vietnam. English is not the main language in my country, so I use AI to help me share my knowledge, experience and lessons learned with learners around the world.)

You type a few words into an AI image tool. The picture comes back with the wrong idea, the wrong style, or text that looks like nonsense.

So you add an adjective, try again, and try again. The tool that was meant to save you hours becomes a slot machine.

The good news: the tool is rarely the problem. The prompt is. And a good prompt is not a talent. It is five layers you can learn in one afternoon.

Meet the 5-Layer Formula. Subject. Action and emotion. Setting and composition. Style and technique. Format and requirements. Each lesson teaches one layer in about three minutes, and every layer runs live in a real AI image tool: the usual short prompt first, then the same idea written with the layer. You see the difference before you are asked to believe it.

You work with real images: a mug, a fox, a cabin, a cup of tea, a sale banner, a cake, a cat by the window. After each lesson you pause and write the layer for a picture you actually need, a product photo, a banner, an illustration or a character.

What happens after the first image? You turn the five layers into a reusable prompt template that works in more than one tool, edit images with follow-up messages by changing one layer at a time, and become your own "image doctor" for distorted hands and misspelled text. The last lesson covers usage rights, real people and disclosing AI images, so you can use your pictures with confidence.

This course is short on purpose, and strict about proof.

After each lesson you take a short quiz where every wrong option explains the mistake. At the end you take two timed practice tests (20 questions each, 14 correct to pass) and hand in a set of five images for five real purposes, each with a "before" and an "after" prompt, graded against a clear rubric. If you pass, you know the formula works for you, not just for the examples.

What you get:

• 11 focused video lessons with English captions
• 27 practice questions with feedback on every answer choice
• 2 full practice tests, a hands-on project of five images, and a scoring guide with grades
• Downloadable resources: the 5-layer checklist, a reusable prompt template, the project brief with rubric, and the score and grade sheet

Who it is not for: anyone looking for advanced image editing software, model training, or video generation. This is about getting the picture you imagined from an AI image tool, in everyday work.

Start the first lesson for free. In about three minutes you will watch one idea turn into two very different pictures.

About how this course was made: it was designed, directed, and reviewed by the instructor, with AI tools assisting in drafting, editing, and production. The narration uses a synthetic voice, the short on-screen presenter at the start of each lesson is an AI-generated avatar built from the instructor's own recorded video, and the background music and some sound effects are synthesized. The images shown in the lessons were generated live with an AI image tool. The teaching method, the examples, and the assessments were chosen and approved by the instructor.