The creator’s playbook The masterclass
ChatGPT Images Mastery Hub

Make the image you have in mind.

From your first prompt to a repeatable creative workflow. Learn to compose, light, edit, and evaluate images through 16 practical lessons grounded in official OpenAI guidance.

16 practical lessons 28 adaptable prompts
A floating orange sphere framed by a monumental travertine arch, reflected in blue water in a sunlit desert gallery
01 / THE ART OF DIRECTION An idea, made intentional.
Warmly illuminated arch and orange sphere above blue water at twilight
Guided audio workshop

See it. Hear it. Try it.

Explore the lighting, choose a mood, and write a direction you can test. Six steps, with narration and visual practice.

What would you like to create?

Explore the sample, then adapt the brief.

01 / Start here

Foundations

Describe your idea with purpose, composition, light, and style.

02 / Build control

Creative control

Direct references, preserve details, and make deliberate revisions.

03 / Develop a workflow

Professional practice

Evaluate results, build a visual system, and prepare a finished set.

Your creative curriculum

Read, practice, inspect, revise. Progress at your own pace.

0 / 16 complete
The masterclass
Foundations
Creative control
Professional practice
Lesson 01 of 16Beginner · 12 min guided practice

Understanding ChatGPT Images

How image creation, editing, references, and limitations work.

First image brief: one focal subject, a quiet setting, and a restrained palette.
Generated course sample

Understanding ChatGPT Images

Cobalt ceramic pitcher and an olive branch on an ivory plaster surface in soft morning light

One possible visual direction.

Original generated still-life example with filled-in template details. Actual output varies and should be reviewed.

This sample illustrates the prompt’s intent. Your subject, references, wording, and results will differ.

Download sample
See the sample’s generation prompt
Create one simple but beautiful editorial still-life photograph for a modern design magazine, landscape3:2. Main subject: a sculptural cobalt-blue glazed ceramic pitcher. Setting: a pale warm plaster tabletop against a softly textured ivory wall. Clear single focal point, one branch of fresh olive leaves resting quietly beside the pitcher, restrained cobalt, ivory and sage palette. Uncluttered composition with breathing room, soft morning window light from camera-left, coherent gentle shadows, rich believable ceramic glaze and handmade surface. Premium professional photograph, refined calm visual hierarchy. No text, logos, watermark or extra objects.
Look closely

Find the subject first. Then name the setting, color choices, and details you would check against the brief.

What you will be able to do

  • Choose generation or editing from a real brief.
  • Identify the parts of an output that require human review.

The core principle

ChatGPT supports creating images from written requests and editing uploaded or generated images. It also supports text and transparent-background requests. Those capabilities are not promises that every instruction will be met. A prompt is a brief; an output is a candidate.

In this course, start with a small still-life brief rather than a complicated campaign. Decide what success means before generating: one cup, a clear focal point, and no text. If you already have a useful composition, try an edit instead of rebuilding it. The learning goal is to distinguish the task from the result and to notice unintended changes.

Course practice: keep a brief, candidate, and review note together. Write what worked, what missed the brief, and your next specific instruction. This hub prepares prompts and tracks practice; generation happens separately in ChatGPT.

  • Choose generation or editing from a real brief.
  • Write one sentence stating the intended use and main subject.
  • The review note separates an observed result from a requested change.

A workflow to try

  1. Write one sentence stating the intended use and main subject.
  2. Choose create for a new scene or edit for an existing image.
  3. List three visible conditions the candidate must satisfy.
  4. Review the output against the conditions and record one next change.

See the difference in the brief

Course-authored example. Use it as an experiment, then judge the image you actually get.

Starting direction

Make a nice portrait.

More specific direction

Create an editorial portrait for a professional profile. Use the attached portrait as the identity reference. Frame head and shoulders against a neutral grey background with soft light from camera-left. Preserve recognizable facial features and natural skin texture. No text.

The revised prompt states purpose, reference role, framing, lighting, and what to preserve.

Your practical assignment

Create a tabletop still life with one cup and no text. Inspect it against three criteria. Then ask for a single background change while keeping the composition. Save both prompts and your review notes.

How to judge your result

  • The chosen task matches the brief.
  • You can point to the main subject and describe one mismatch.
  • The review note separates an observed result from a requested change.
Common mistakes to watch for
  • Calling the first candidate finished because it looks attractive.
  • Confusing image analysis with image generation.
  • Using a private reference without permission.

Check your understanding

You have a useful product composition but need a different background. What is the most appropriate first task?

Your lesson notebook