ChatGPT · Module 01 of 8

Basics

Get productive with ChatGPT fast — the core mental model and first prompts.

2 lessons8 exercises
Topics & techniques

Getting started

Accounts, models, and your first useful prompt.

4 cards →
Knowledge card
Build Your First Prompt From Context, Goal, and Format
A three-part template for any prompt: state the context you're working in, the specific goal you want, and the format the answer should take. It's the fastest fix for disappointing first answers.
L1 · Basic prompts
Knowledge card
Pick the Right ChatGPT Model for the Job
Choosing which ChatGPT model runs your prompt — the fast general model for most work, a reasoning model when the task needs real step-by-step thinking, and deep research for long sourced investigations.
L1 · Basic prompts
Best practice
Start a New Chat When the Topic Changes
A context-hygiene habit: when you switch to an unrelated task, open a fresh chat instead of continuing the old thread, so stale context doesn't colour the new answer.
L1 · Basic prompts
Comparison
Free, Plus, or Pro: Which ChatGPT Tier You Actually Need
A plain rule for choosing a tier: the free tier covers most learning and everyday tasks; paid tiers pay off when you keep hitting usage limits or lean on the strongest reasoning models and Deep Research.
L1 · Basic prompts

Custom instructions

Set persistent context so you stop repeating yourself.

4 cards →
Knowledge card
Fill Both Boxes: Who You Are and How to Answer
A guide to the two custom-instruction fields: the first tells ChatGPT who you are and what you're doing; the second tells it how you want answers shaped. Both apply to every new chat automatically.
L1 · Basic prompts
Comparison
Custom Instructions vs Memory vs Projects: What Goes Where
A routing guide for the three places ChatGPT stores context: global preferences go in custom instructions, project-specific rules and files go in a Project, and durable facts it learns go in Memory — with one-off detail staying in the prompt.
L2 · Context engineering
Best practice
Set Custom Instructions Once, Stop Repeating Yourself
Custom instructions are two standing fields — who you are and how you want answers — that ChatGPT applies to every new chat automatically.
L1 · Basic prompts
Best practice
Tune Custom Instructions From Answers You Didn't Like
A maintenance habit: treat every correction you find yourself repeating ('shorter', 'no bullet points', 'British spelling') as a signal to add a permanent custom instruction, so you fix it once instead of every chat.
L2 · Context engineering
ChatGPT basics, explained

Start with the model picker, not the prompt

The first thing to understand about ChatGPT is that it is not one model but several, and the picker at the top of the screen matters more than most beginners realise. The fast default handles everyday writing, explaining, and quick questions at speed. The reasoning models are slower but think through multi-step problems, maths, and planning before they answer. Choosing well is the difference between a fluent wrong answer and a correct one. If you only learn one thing from the basics, make it this: match the model to the task before you type.

The habit that fixes most bad answers

Beginners tend to type a few words and hope. Better results come from giving ChatGPT three things every time — context (who you are and what you are doing), a clear goal (the exact outcome you want), and a format (length, structure, tone). 'Explain APIs' is vague; 'Explain REST APIs to a designer with no backend experience, in under 200 words, using one real analogy' is not. Once you settle into a way of working, set your custom instructions so ChatGPT carries that context into every new chat and you stop repeating yourself.

Where to go from here

The basics split into two moves worth mastering early: getting started (accounts, the interface, and your first genuinely useful prompt) and custom instructions (persistent context that makes every future chat sharper). Work through the knowledge cards below in order — each one is a small, tested habit rather than a wall of theory.

Frequently asked
How do I start using ChatGPT?

Create a free OpenAI account, open the chat box, and start with a clear request — say who you are, what you want, and the format you need back. You do not need to install anything, and the free tier is enough to learn on. The getting-started cards below walk through your first useful prompts step by step.

Do I need to pay for ChatGPT?

No. The free tier gives you a capable default model and is plenty for learning and everyday tasks. Paid plans add higher limits, the strongest reasoning models, and extra features — worth it once you rely on ChatGPT daily or need the reasoning models regularly, but not necessary to get started.

What is the most important beginner habit?

Give context, a goal, and a format in every prompt, and pick the right model for the task. Those two habits fix the majority of disappointing answers. Setting your custom instructions once so ChatGPT remembers your role and preferences is the natural next step.

Why do I get generic answers from ChatGPT?

Almost always because the prompt was generic. Without context ChatGPT writes for an average reader and an average purpose. Tell it who the answer is for, what you are trying to achieve, and how you want it structured, and the output sharpens immediately.

Test yourself

Do you know the ChatGPT basics?

Six questions on the fundamentals — building a prompt that works, picking the right model, and setting custom instructions so you stop repeating yourself. No sign-up, instant score, and you can share how you did.

  1. Q1 What are the three parts of a prompt that reliably beats a one-liner?
  2. Q2 When is a reasoning model the wrong choice?
  3. Q3 You switch to a completely unrelated task. What's the right move?
  4. Q4 When does upgrading from the free tier actually pay off?
  5. Q5 What do the two custom-instruction fields do?
  6. Q6 Project-specific rules belong in…
0 / 6 answered