Fast models vs. thinking models — Foundations
Some models think first
Reasoning models (often labeled Thinking, Extended thinking, Deep Think or a high "effort" setting) work through a problem privately — breaking it into steps, checking their work, backtracking — before answering.
The result: much better at math, logic, planning, analysis and code. The cost: you wait longer.
📝 Read the full lesson notes
Some models think first
Reasoning models (often labeled Thinking, Extended thinking, Deep Think or a high "effort" setting) work through a problem privately — breaking it into steps, checking their work, backtracking — before answering.
The result: much better at math, logic, planning, analysis and code. The cost: you wait longer.
Most apps give you the choice
ChatGPT, Claude, Gemini and Grok all offer a fast mode and a thinking mode — via a model picker, a toggle, or an automatic router that decides for you.
Rule of thumb: if a smart friend would need a notepad to answer it, turn thinking on.
Try it: race the modes
A bat and a ball cost $1.10 in total. The bat costs $1.00 more than the ball. How much does the ball cost? Explain briefly.
Part of 🧠 Foundations in the free AI Bootcamp.