Learn DALL·E in 5 minutes.
DALL·E is OpenAI's image generator, used mostly from inside ChatGPT with plain conversational instructions. Here you will practice describing scenes precisely, requesting aspect ratios, and using the select tool to edit one region without regenerating the rest.
Zeteo is taking you through this one. Your docent for this room. Free, and it runs right here.
Lesson not loading? Open it in a new tab. It is the same free lesson either way.
New image, or new sentence?
In ChatGPT, generation and revision are the same conversation. The skill is knowing when to refine with a follow-up, when to select and edit, and when to start clean.
Describe the scene
You write what you want in normal language and ChatGPT expands it into a detailed image prompt. Subject, setting, style, and lighting all belong in your description if you care about them.
Revise in conversation
Say what to change, the same scene but at night, and DALL·E revises. Small course corrections work well; stacking many revisions slowly degrades the image.
Edit a region
The select tool lets you brush over part of a generated image and describe a change for just that area. Use it when nine tenths of the image is already right.
What you will do in the lesson
You answer and try things. Nobody demonstrates at you for five minutes.
- Say what you actually want. You turn a vague request into a specific one, naming subject, style, and mood, and compare what comes back.
- Control the frame. You request wide, square, and tall versions and learn how aspect ratio changes composition, not just shape.
- Revise one change at a time. You practice follow-ups that alter a single element, and see why bundled changes come back scrambled.
- Fix a region with select. You brush over one part of an image and rewrite only that, keeping everything else stable.
- Recognize a dead end. You learn to spot when a conversation has drifted too far and a fresh prompt beats another revision.
Before you start
Is DALL·E a separate app?
For most people it lives inside ChatGPT: you ask for an image the way you ask for anything else. There is no separate interface to learn, which is exactly why prompt precision is the whole skill.
Do I have to install anything to take this lesson?
No. The lesson runs right here in the page, free, with no signup. If you use ChatGPT already, you can apply what you practice immediately; image generation is available with a free way to try it.
Can it render readable text in images?
Short text often works, but longer or smaller text still comes out mangled at times. For designs where typography must be exact, a tool built for text rendering such as Ideogram is the better fit.
DALL·E or Midjourney?
DALL·E is stronger at following complicated instructions and editing conversationally; Midjourney is stronger at raw visual style and fine parameter control. Many people draft with DALL·E for accuracy and switch to Midjourney when looks are the priority.
Other tools people learn next
Same format, same five minutes, still free.
Learners from these organizations are already learning on 
