AI Tokenomics in a Minute

AI Tokenomics

Tokens, caching, context growth, pricing mechanics, and cost per engineering outcome.

Central question: What engineering outcome did those tokens purchase?

• Tokens

  What tokens are in LLM APIs, how input and output differ, and why token counts drive coding-agent cost.

• Prompt Caching

  How prompt caching can reduce repeated-input cost, and what it does not guarantee.

• Cost Per Task

  How to estimate what a coding-agent task costs in tokens and dollars—with assumptions visible.

• Input vs Output Tokens

  Why input and output tokens are priced and constrained differently in coding-agent workloads.

• Context Windows

  What a context window is, how it bounds agent sessions, and what a large window does not buy you.

• Context Growth

  Why agent sessions get more expensive over time as history and tool output accumulate.

• Retries and Agent Loops

  How retries and tool loops multiply token spend—often without multiplying value.

• Tool Calls

  How agent tool use (read, search, shell, tests) drives tokens, latency, and cost beyond the model completion itself.