API
Also called application programming interface, AI API
The way one program talks to another. AI companies sell model access through APIs, so other apps can build their features on top.
An API is a doorway built for software rather than people. A program sends a request in an agreed format and gets an answer back in an agreed format, with no screen or buttons involved.
For AI, this is how most of the world uses the big models without visiting ChatGPT or Claude. The writing helper in your email app, the chatbot on a bank's website and the summary button in a note-taking app are usually sending text to one of a handful of model makers through their APIs and showing you what comes back.
Developers pay per use, measured in tokens. That's why apps put limits on their AI features, and why the same model can behave a little differently inside another app: the app writes its own instructions around your request.
For example
A recipe app's "turn this into a shopping list" button sends your recipe to an AI model through an API and shows you the reply.
Put it to use
Where you'll meet API in ChatGPT, Gemini and Claude, how to use it there, and the mix-up most people make.
Daily Bytes Plus is in the Android app.