Autocompletion

Autocompletion, or autocomplete, is a software feature that predicts the rest of a word or phrase as a user types, and often inserts the predicted text automatically or upon selection. It is used to speed up input, reduce keystrokes, and improve accuracy across contexts such as text editors, integrated development environments, web browsers, and command-line interfaces.

Core mechanisms include dictionaries of terms, probabilistic models, and machine-learning systems that rank candidate completions by likelihood given the current context. Traditional approaches rely on static dictionaries or n-gram models; more recent systems employ neural networks, transformer-based models, and context-sensitive coding models. Autocomplete can draw on general language data, domain-specific corpora, or user history and project context to tailor suggestions.

Common interfaces present one or more candidate completions in a dropdown or inline hint. Users can accept a suggestion with a keystroke (often Tab or Enter) or dismiss it. Completion can be triggered explicitly (Ctrl-Space) or automatically as characters are typed. In code editors, names, methods, and parameters are suggested, sometimes with signature and type information; in general text, suggestions may include words, phrases, or templated snippets.

Advantages include faster input and consistency; drawbacks include privacy concerns, potential bias or inappropriate recommendations, and reliance on imperfect models. Settings often allow disabling autocompletion or tuning its behavior.