Free prompt engineering course
This free, self-paced prompt engineering course teaches you to write prompts that get reliable results in ChatGPT, Gemini and Claude. It has six short lessons, each pairing a guide with a hands-on exercise and a quick self-check, and takes about two hours. There is no sign-up and no fee.
By the ThinkWithPrompt editorial team at Synottic · 6 lessons · about 2 hours · free, no sign-up
Start lesson 1Open the prompt generator
What will you learn in this prompt engineering course?
- Explain what a prompt is and how an AI model uses it
- Write prompts with a role, context, task and format
- Choose between the RTF, RCTF and RCTFVV frameworks
- Use examples (few-shot) and step-by-step reasoning (chain-of-thought)
- Write for a specific audience with CO-STAR
- Build, test and save reusable prompts for your own work
Lesson 1: What a prompt is and how AI reads it
About 15 minutes
Read: What is a prompt in AI?.
Practice: Pick a task you did this week. Write the one-line prompt you would normally use, then add the context only you know (audience, goal, constraints). Run both in your AI assistant and compare the answers.
Check yourself: What is the difference between a prompt and a prompt template?
Lesson 2: The four parts of a good prompt
About 20 minutes
Read: How to write a good AI prompt, then The RCTF framework.
Practice: Rewrite your Lesson 1 prompt with all four parts: Role, Context, Task and Format. Use the prompt generator to check your structure.
Check yourself: Which part of your prompt removed the biggest guess for the AI?
Lesson 3: Choosing the right structure
About 15 minutes
Read: RTF vs RCTF vs RCTFVV.
Practice: Take one quick task and one high-stakes task from your work. Write the first in RTF and the second in RCTFVV, with the verification and validation steps.
Check yourself: When is the extra length of RCTFVV worth it?
Lesson 4: Showing examples and asking for reasoning
About 20 minutes
Read: Few-shot vs chain-of-thought prompting.
Practice: Write a few-shot prompt with two examples for a task that needs a consistent format (for example, classifying feedback). Then write a chain-of-thought prompt for a decision you need to make.
Check yourself: Which technique improves consistency, and which improves reasoning?
Lesson 5: Writing for an audience
About 15 minutes
Read: The CO-STAR prompt framework.
Practice: Write the same announcement twice with CO-STAR: once for senior leaders and once for new joiners. Compare how style, tone and audience change the output.
Check yourself: What do the S, T and A in CO-STAR add that RCTF leaves implicit?
Lesson 6: Prompt engineering in practice
About 30 minutes
Read: What is prompt engineering?.
Practice: Final exercise: choose three tasks you repeat every week. For each, find a close match in the library or build one with the generator, run it, refine it until the output is consistently good, and save it for reuse.
Check yourself: What would you change in a prompt that gives a good answer only some of the time?
Frequently asked questions
Is this prompt engineering course really free?
Yes. All six lessons, the guides and the exercises are free, with no sign-up. You only need access to an AI assistant such as ChatGPT, Gemini or Claude, and their free versions are enough.
Do I get a certificate?
No. This self-paced course doesn’t award a certificate. Organisations that want structured, instructor-led training for their teams can look at the Prompt Engineering programme by the Synottic AI Institute, described above.
Do I need to know coding?
No. The course is about clear, structured instructions for everyday AI assistants. No programming is required.
How long does the course take?
About two hours: roughly 15–30 minutes per lesson including the exercise. You can take it at your own pace and come back to any lesson.
Which AI tool should I use for the exercises?
Any of them. The techniques work in ChatGPT, Gemini, Claude and other assistants. Try the same prompt in two tools to see how they differ.