Course plan › Track A · Free › Module 1
Meet Claude
Five short lessons. You learn what Claude is, how cost and usage are counted, which model to choose and how to avoid the most common mistakes.
- Explain what a token is
- Work out a simple cost
- Choose a model for a job
- Know what never to type in
What is Claude?
4 min
Claude is an AI assistant from Anthropic. You write, Claude answers. Claude can be wrong, so always check what matters.
Explanation
Claude is what is called a language model. It is trained on a large amount of text and can write, summarize, translate, explain and help you think. You reach Claude through the web, apps for desktop and mobile, and tools for developers.
According to Anthropic's pricing page, the free plan already includes things like chat on web, desktop and mobile, web search, creating files and running code, memory across conversations, connecting apps, and artifacts (finished documents and small tools).
An AI can write something that sounds completely right but is not. This applies to numbers, sources, laws and quotes. Ask for sources and check them yourself when it matters.
Think of one thing you would like help with this week. Write it on a note. You will use it in the next lesson.
Your first conversation
5 min
Create an account, write a clear question, and ask Claude to ask follow-up questions. You can also steer how the answer looks.
How to do it
- Go to claude.ai or download the app and create an account. You create the account yourself. Never share your password with anyone, not even with Claude.
- Type your question in the box at the bottom.
- Read the answer. Reply with a new message if you want to change something.
Three good first prompts
Feel free to copy and try them:
Explain what a token is as if I were twelve years old. Five sentences at most.
Help me write a friendly email to my landlord about a dripping kitchen tap. Ask me first if there is anything you need to know.
Always answer briefly: one heading and five bullet points at most. Confirm with one word that you understood.
They say what you want, for whom, and how long the answer should be. The second lets Claude ask first. The third controls the format. More on this in module 2.
Try one of the prompts above, or your note from lesson 1. Then ask Claude to make the answer shorter or simpler.
Tokens: the fuel of AI
6 min
A token is a small piece of text. Claude reads and writes in tokens. The more text that is sent, the more is used. What Claude writes costs more per token than what you send in.
Explanation
A word can be one token, several tokens or part of a token. How many tokens the same text becomes depends on language, content and which model it is. Anthropic states that Claude 4.7 and later models use a newer tokenizer that produces roughly 30 percent more tokens for the same text than earlier models. So you cannot count exactly in your head, but you can still keep a reasonable overview.
Two ways to pay
| Subscription (Free, Pro, Max) | API (for developers) | |
|---|---|---|
| Payment | Fixed monthly price. Pro is USD 17/month with annual billing (USD 20 if billed monthly). Max starts at USD 100/month. | You pay per token used. |
| Limits | Usage limits that reset on a rolling five-hour window, plus weekly limits on paid plans. There is no fixed number of messages. | You pay for what you use. You can set spending caps yourself. |
All prices below are in US dollars and exclude VAT.
A small calculation (API)
Sonnet 5.5 costs USD 2 per million input tokens and USD 10 per million output tokens. If you send 10,000 tokens and get 2,000 tokens back:
- Input: 10,000 / 1,000,000 × USD 2 = USD 0.02
- Output: 2,000 / 1,000,000 × USD 10 = USD 0.02
- Total: USD 0.04
The working memory
All four current models have a context window of 1 million tokens and can write up to 128,000 tokens in one answer. The context window is like a desk. Whatever is on it, Claude can see in that conversation.
- Start a new conversation when you change topic. Old text otherwise comes along.
- Be specific. A clear question produces less clutter.
- Use a smaller model for simple jobs (next lesson).
Anthropic gives the same advice in its documentation on Claude Code costs. You will go deeper in module 4.
Four models in one minute
6 min
Claude comes in four sizes. Small ones are fast and cheap. Large ones think longer and cost more. Pick the smallest one that does the job.
| Model | Best for | Speed | Input / output |
|---|---|---|---|
| Haiku 5.5 | Fast, high-volume jobs: classification, data extraction and routing | Fastest | From 0.10 / 0.50* |
| Sonnet 5.5 | Best combination of speed and intelligence | Fast | 2 / 10 |
| Opus 5.5 | Long-running coding and knowledge work | Moderate | 4 / 20 |
| Fable 5.1 | Demanding reasoning and long agent tasks | Slower | 10 / 50 |
*Haiku 5.5 costs more when a request is over 100,000 tokens (0.50 / 2.50).
Which one do I get in the app?
According to Anthropic's pricing page: Free gives Sonnet and Haiku. Pro adds Opus, and Fable through extra usage credits. Max gives Fable with up to 50 percent of the weekly limit. The selection can change, so check the model picker in your app.
Sonnet is described as the best balance of speed and intelligence. Anthropic's API documentation also suggests Opus 5.5 as the starting point for most workloads when you build services. For everyday chat use, it is reasonable to start with a smaller model and move up if the result is not enough.
Ask the same question to two models (if your plan has more than one). Compare the answers and how fast you got them.
Safety first
4 min
Do not enter secret or sensitive information. Check your privacy settings. Review answers that matter.
Do not enter
- Passwords, codes and API keys
- National ID numbers, bank cards and account details
- Other people's personal data, unless you know you are allowed to (GDPR)
- Health data and other sensitive details about yourself or others
- Your employer's confidential material, unless you have an agreement that allows it
Check your settings
Anthropic's pricing overview lists model training as an opt-out choice for Free, Pro and Max. For Team and Enterprise, no training is listed as the default. Open Settings in your app and check what applies to you.
Review the answers
Check numbers, laws and sources. Ask Claude for sources and open them yourself. In module 6 you will learn how to fact-check systematically.
Check yourself
Think first, then open the answer.
1. What is a token?
2. Which costs more per token, input or output?
3. Which model suits simple sorting in large volume?
4. May you enter passwords or ID numbers in a chat?
5. Why is it good to start a new conversation when you change topic?
Well done!
You have completed module 1. Next is module 2, "Talk to Claude".
Sources (checked 2026-10-13)
- Anthropic: Models overview
- Anthropic: Pricing (API)
- Claude: Plans & pricing
- Claude Code: Manage costs effectively
Prices and models can change. Always check the official pages.