What Is a Vector in AI?

A vector is a list of numbers that represents data as a point in a multi-dimensional space. In AI, vectors are how information, from text to images, is encoded numerically so a model can work with it, and the distance between two vectors gives a way to measure how similar the underlying data is.

How a vector works

A vector is simply an ordered list of numbers, and each number places the data a little further along one dimension. With enough dimensions, a vector can pin data to a precise point in a large space. Models represent almost everything this way internally, because math over numbers is something computers do well, while raw text or pixels is not.

The useful property is distance. When related data is placed close together and unrelated data far apart, the gap between two vectors becomes a measure of similarity. Comparing vectors is how a system decides that two sentences mean nearly the same thing or that two images look alike, without understanding them the way a person would.

Why vectors matter for AI

Vectors matter because they are the common language AI uses to represent and compare information. An embedding, the meaning-carrying representation behind semantic search and retrieval, is a vector, and a vector database exists specifically to store and search them at scale. Almost any time an AI system judges similarity or works with meaning, vectors are doing the work underneath. Understanding them makes related ideas like embeddings and similarity search much clearer.