Learn

LLM API Costs & Economics, Explained

Every article here pairs with a calculator on this site. LLM pricing changes frequently — these articles explain the underlying cost structure so you can make provider and architecture decisions that hold up as pricing evolves.

Token Cost

LLM cost per million tokens explained: input, output, and cachingInput tokens, output tokens, and cached tokens price at different rates — and output tokens are typically 3-5× more expensive. Here's the model-agnostic framework for understanding LLM cost at scale.

Prompt Caching

Prompt caching ROI by use case: when it pays and when it doesn'tPrompt caching breaks even at ~60% cache hit rate for most workloads. Customer support and RAG retrieval are natural fits; one-off generation rarely justifies the write cost. Here's the use-case breakdown.