No specific prices, since they change every few months. What this reference lesson gives you is the structure of AI costs so you can evaluate any option in front of you: the API token as the base unit, paying per token versus subscription plans, and why plans tend to be the better deal. I share my own pattern for mixing daily-driver models with a heavier one for the hardest work, and the progression I recommend: start on one provider's lowest paid plan, graduate as you hit limits, and stack plans as you start mixing. It closes on the point the whole course is built on: nothing here depends on one model or provider.
What's covered:
- The API token as the base unit of AI cost
- Paying per token vs subscription plans
- Mixing and matching frontier models: daily drivers and a heavy hitter
- The progression: start small, graduate on limits, stack plans when you need to
- Why several hundred dollars a month can be rational for a professional
- Portability: shifting spend as models and tools change