Cost Arithmetic · ai-skills

Reference: Tokens, Cost and Latency · Beginner to Intermediate

4topics
4lessons

Hosted AI calls are billed from token counts. This topic builds cost arithmetic from EXAMPLE prices through two engineering jobs to monthly totals.

Term	Meaning
Input tokens	Tokens sent in: instructions, context and data.
Output tokens	Tokens received back: answers, summaries and reviews.
EXAMPLE price	A stated classroom rate per 1M tokens, never a real vendor price.
Cost formula	Cost equals input tokens times input price plus output tokens times output price.

Cost as the sum of an input part and an output part.

Input and output parts add to one total, each in EXAMPLE units.

Lesson	What it covers
The cost formula	Price units, the input-output split, and one fully worked call.
A code-review job	Tokens per review and the step to daily totals.
A nightly regression summary	Tokens per night and the 30-night monthly projection.
Sensitivity and a calculator	What doubling the context does, plus a runnable cost calculator.

Check yourself A call uses 8000 input tokens and 500 output tokens at the EXAMPLE prices of 3 and 12 units per 1M. Which part dominates, and what is the total?

The input part dominates at 0.024 units against 0.006 for output. The total is 0.030 units.

Up nextThe Cost Formula

Practice this lesson Report an error