Who it's for
For people who already use ChatGPT, Claude or Gemini and are ready to read short Python snippets. The first lesson works in a chat app up to a point, and the code in the others needs Python 3 and an API key from at least one provider.
Lessons
- 1What are reasoning models? AI that thinks before it answersA reasoning model writes hidden thinking tokens before it answers. See how OpenAI, Claude, Gemini and DeepSeek-R1 do it, and read the token count yourself.
- 2How to prompt reasoning models: write the goal and leave out the stepsHow to prompt reasoning models such as GPT-5, Claude and DeepSeek-R1: state the goal, limits and output format, drop step-by-step cues and fix clashing rules.
- 3Reasoning effort and thinking budgets: how to set them and what they costSet reasoning effort on OpenAI, Claude and Gemini: the parameters, token caps that prevent empty replies, what each level costs in tokens and how to pick one.
- 4When to use a reasoning model: which tasks gain and which do notWhich tasks gain from a reasoning model and which do not: OpenAI's planner and workhorse advice, DeepSeek's own scores and research on overthinking.
Frequently asked questions
Is this reasoning models course free?
Yes. Every lesson is free to read. The code steps need an API key from OpenAI, Anthropic or Google, and the providers charge for API use under their own terms.
Do I need to know how to code?
You should be able to run a short Python script. The examples are small and every lesson explains the lines that matter. A chat app is enough for the first half of lesson 1 and for rewriting a prompt in lesson 2.
Which models does the course cover?
OpenAI's reasoning models, Claude with extended thinking, Gemini with its thinking levels and DeepSeek-R1. The vendors rename models often, so each lesson names the model its source used and tells you to check the current docs.
How is this different from the chain-of-thought lesson?
The chain-of-thought lesson in the advanced prompt engineering course covers how to ask a model to reason. This course starts one step later, with models that already reason on their own, and covers what you change in your prompt, your settings and your choice of model.
What should I learn after this course?
Prompt optimization is next. It teaches AI to draft, improve and test the prompts themselves, and it builds on the habit of testing on your own samples that lesson 4 here starts.