API

What it means. An API (application programming interface) is a defined way for one piece of software to ask another for something and get a structured answer back. When your app uses an AI model, it talks to that model through an API.

Why it matters. Almost every AI feature you add depends on an API call to a provider like OpenAI or Anthropic. That call has rules: an address, a key, a request format, limits, and a cost. Understanding the API is understanding what your feature actually does.

What beginners often misunderstand. An API is not the AI itself, and it is not free or unlimited. Each call leaves your system, runs on someone else’s servers, and can fail, slow down, or change.

How it shows up in real projects. API keys get exposed by accident, rate limits get hit at launch, response formats change and break features, and costs climb when call volume is not tracked. Most AI reliability problems start at the API boundary.